Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Characterization of the Drosophila Group Ortholog to the Amino-Terminus of the Alpha-Thalassemia and Mental Retardation X-Linked (ATRX) Vertebrate Protein

  • Brenda López-Falcón ,

    Contributed equally to this work with: Brenda López-Falcón, Silvia Meyer-Nava

    Affiliation Departamento de Genética del Desarrollo y Fisiología Molecular, Instituto de Biotecnología, Universidad Nacional Autónoma de México, Cuernavaca, Morelos, México

  • Silvia Meyer-Nava ,

    Contributed equally to this work with: Brenda López-Falcón, Silvia Meyer-Nava

    Affiliation Departamento de Genética del Desarrollo y Fisiología Molecular, Instituto de Biotecnología, Universidad Nacional Autónoma de México, Cuernavaca, Morelos, México

  • Benjamín Hernández-Rodríguez,

    Affiliation Departamento de Genética del Desarrollo y Fisiología Molecular, Instituto de Biotecnología, Universidad Nacional Autónoma de México, Cuernavaca, Morelos, México

  • Adam Campos,

    Affiliation Departamento de Medicina Molecular y Bioprocesos, Instituto de Biotecnología, Universidad Nacional Autónoma de México, Cuernavaca, Morelos, México

  • Daniel Montero,

    Affiliation Departamento de Genética del Desarrollo y Fisiología Molecular, Instituto de Biotecnología, Universidad Nacional Autónoma de México, Cuernavaca, Morelos, México

  • Enrique Rudiño,

    Affiliation Departamento de Medicina Molecular y Bioprocesos, Instituto de Biotecnología, Universidad Nacional Autónoma de México, Cuernavaca, Morelos, México

  • Martha Vázquez,

    Affiliation Departamento de Genética del Desarrollo y Fisiología Molecular, Instituto de Biotecnología, Universidad Nacional Autónoma de México, Cuernavaca, Morelos, México

  • Mario Zurita ,

    vvaladez@ibt.unam.mx (VVG); marioz@ibt.unam.mx (MZ)

    Affiliation Departamento de Genética del Desarrollo y Fisiología Molecular, Instituto de Biotecnología, Universidad Nacional Autónoma de México, Cuernavaca, Morelos, México

  • Viviana Valadez-Graham

    vvaladez@ibt.unam.mx (VVG); marioz@ibt.unam.mx (MZ)

    Affiliation Departamento de Genética del Desarrollo y Fisiología Molecular, Instituto de Biotecnología, Universidad Nacional Autónoma de México, Cuernavaca, Morelos, México

Characterization of the Drosophila Group Ortholog to the Amino-Terminus of the Alpha-Thalassemia and Mental Retardation X-Linked (ATRX) Vertebrate Protein

  • Brenda López-Falcón, 
  • Silvia Meyer-Nava, 
  • Benjamín Hernández-Rodríguez, 
  • Adam Campos, 
  • Daniel Montero, 
  • Enrique Rudiño, 
  • Martha Vázquez, 
  • Mario Zurita, 
  • Viviana Valadez-Graham
PLOS
x

Correction

10 Feb 2016: López-Falcón B, Meyer-Nava S, Hernández-Rodríguez B, Campos A, Montero D, et al. (2016) Correction: Characterization of the Drosophila Group Ortholog to the Amino-Terminus of the Alpha-Thalassemia and Mental Retardation X-Linked (ATRX) Vertebrate Protein. PLOS ONE 11(2): e0149367. https://doi.org/10.1371/journal.pone.0149367 View correction

Abstract

The human ATRX gene encodes hATRX, a chromatin-remodeling protein harboring an helicase/ATPase and ADD domains. The ADD domain has two zinc fingers that bind to histone tails and mediate hATRX binding to chromatin. dAtrx, the putative ATRX homolog in Drosophila melanogaster, has a conserved helicase/ATPase domain but lacks the ADD domain. A bioinformatic search of the Drosophila genome using the human ADD sequence allowed us to identify the CG8290 annotated gene, which encodes three ADD harboring- isoforms generated by alternative splicing. This Drosophila ADD domain is highly similar in structure and in the amino acids which mediate the histone tail contacts to the ADD domain of hATRX as shown by 3D modeling. Very recently the CG8290 annotated gene has been named dadd1. We show through pull-down and CoIP assays that the products of the dadd1 gene interact physically with dAtrxL and HP1a and all of them mainly co-localize in the chromocenter, although euchromatic localization can also be observed through the chromosome arms. We confirm through ChIP analyses that these proteins are present in vivo in the same heterochromatic regions. The three isoforms are expressed throughout development. Flies carrying transheterozygous combinations of the dadd1 and atrx alleles are semi-viable and have different phenotypes including the appearance of melanotic masses. Interestingly, the dAdd1-b and c isoforms have extra domains, such as MADF, which suggest newly acquired functions of these proteins. These results strongly support that, in Drosophila, the atrx gene diverged and that the dadd1-encoded proteins participate with dAtrx in some cellular functions such as heterochromatin maintenance.

Introduction

The human ATRX gene (hATRX) was described approximately 20 years ago as the main gene mutated in ATRX syndrome (Alpha-Thalassemia with mental Retardation X-related). ATRX is localized on the X chromosome in position Xq13.1–q21.1. Individuals with mutations in this gene present several phenotypic charateristics that may include mental retardation, craniofacial and urogenital deformities, psychomotor failure and alpha-thalassemia [1]. Since its description, there have been important advances in the characterization of the molecular functions of the protein encoded by this gene. In humans, there are mainly two isoforms named hATRX (289 kDa) and hATRXt (t, from truncated, 200 kDa) that are encoded by this gene [2]. Both proteins contain an amino-terminal domain which is composed of PHD and GATA-like zinc fingers, named ADD after the three proteins that contain this domain (ATRX, DNMT3b and DNMT3L). It was recently demonstrated through different in vitro and in vivo approaches that this domain recognizes the combination of K9me3 and unmethylated K4 residues of the histone H3 tail [3]. This domain directs the protein mainly to pericentric heterochromatin [4]. Mutations described in patients afflicted with the syndrome mainly affect the important amino acids that form the "pocket" of the ADD domain for the histone H3 tail recognition. The hATRX protein additionally has a helicase/ATPase domain, which classifies it as a member of the SNF2 subfamily of chromatin remodelers [5]. The hATRX SNF2 domain has in vitro ATPase activity, which can be stimulated both by DNA and nucleosomes [6]. About 50% of the mutations described in patients fall in the ADD domain, whereas the other 50% affect the SNF2 helicase/ATPAse and other protein domains [7]. hATRX, as many chromatin remodelers, has been identified as a component of a complex that includes the histone variant H3.3 chaperone DAXX (Death domain Associated protein). This particular histone variant is incorporated at different chromatin regions, such as promoters, enhancers and heterochromatic regions, and it has been proposed to have dual functions in promoting both an active chromatin state and the maintenance of heterochromatin [8]. hATRX ATPase activity is important for incorporation of the histone variant H3.3 by the chaperone DAXX into specific regions of the chromosomes, such as telomeres and pericentric heterochromatin [9].

Genome-wide studies have identified hATRX as a protein that is able to bind to DNA regions that can acquire a G4 structure conformation, such as telomeres and repetitive G-rich regions [9]; however, the important domain that mediates this interaction it is not yet known.

It is clear though that both the ADD and helicase/ATPase domains play crucial roles during development [1]. The ATRX gene is highly conserved through eukaryotic evolution, but in invertebrates and particularly in Drosophila, it encodes proteins that lack the ADD N-terminal domain, i.e., the dAtrxL and dAtrxs isoforms are encoded by the same gene but lack an ADD domain. The dAtrxL isoform interacts with HP1a and is localized to heterochromatin, whereas the dAtrxs isoform is localized to euchromatin [10].

We decided to determine if there was a gene in Drosophila that could encode proteins with the conserved the amino-terminus of the vertebrate ATRX. Performing in silico analyses, we demonstrate that the annotated gene CG8290, recently named dadd1 [11], encodes three proteins with a conserved ADD domain that physically interact with dAtrxL. Using a genetic approach, we found that these proteins have important functions during Drosophila development and that they cooperate with dAtrxL in certain functions. The evidence leads us to propose that the dAdd1 proteins are orthologs of the amino-terminus of the ATRX protein in vertebrates.

Materials and Methods

Ethics statement

All animal handling was approved by the Instituto de Biotecnología Bioethics Comittee, Permit Number 344 (2011/02/10), which follows NOM-062 animal welfare mexican law. All efforts were made to minimize animal suffering. Animals were sacrificed by CO2 euthanasia.

Protein domain structure illustration

The domain organization of dAdd1 isoforms (Fig. 1C) and the representation of the fragments assayed by pull-down were performed using the DOG 1.0 Illustrator of Protein Domain Structures [12]. For gene representation we used FancyGene [13].

thumbnail
Figure 1. The ADD domain of the hATRX protein is conserved in the dAdd1 proteins of Drosophila.

A. ADD domain prediction in the protein sequence alignment of hATRX and dAdd1 proteins. Conserved cysteines are shown as yellow letters within a cyan box. The amino acids involved in H3K9me3 and H3K4me0 recognition are marked by green and red boxes respectively. The GATA-like C2-C2 zinc finger and the C4-C4 imperfect PHD are marked by green and red boxes respectively. B. Ribbon representation of ADD domain of hATRX on red (left) and dAdd on green (center) CPH models-3.0 server was used to create the model of the ADD domain of dAdd1. Structural superposition of ADD domains of hATRX and dAdd1 (right). C. Domain organization of dAdd1 (a-c) isoforms. All of them have an N-terminal ADD domain. A C-terminal MADF domain is present in two (dAdd1-b, and dAdd1-c isoforms) copies. D. Maximum Likelihood Phylogenetic Analysis of the ADD domain and the corresponding Protein Domain Architecture information of the containing proteins as computed by PhyML [16] and ScanProsite [19], respectively. The numbers shown represent bootstrap values. Note that the ADD harboring protein underwent a gene fission event during the evolution of insects. Also note that the homologous proteins within the Drosophila genus have acquired a tandem MADF domain during its divergence from other insects. This domain is likely to be functional given its conservation within the genus. For the parameters used, refer to Material and Methods.

https://doi.org/10.1371/journal.pone.0113182.g001

Alignments and phylogenetic inference analyses

Multiple alignments were performed with CLUSTALX2 2.1 [14] and the parameters for the Phylogenetic Inferences were used as estimated by ProtTest 2.4 program [15] for selection of models of protein evolution. The Maximum Likelihood Phylogenetic Analysis was computed by PhyML 3.0 [16] with the parameters: Substitution model: WAG (ADD) & LG (Helicase/ATPase) [17]; Bootstrap: 1000; Proportion of invariable sites: 0.12 (Both); Gamma shape parameter: 2.15 (ADD) & 0.81 (Helicase/ATPase). The tree was edited using Interactive Tree Of Life (iTOL) v2 [18] with Protein Domain Architecture information of the containing proteins as predicted by ScanProsite [19].

Protein structure homology model

In an attempt to determine a protein-modeling of ADD domain (101 amino acid residues), we generated by SWISS-MODEL [20], a three-dimensional structural model of a protein target, based on identity sequence related with structures deposited in the Protein Data Bank. Basically the steps used in homology modeling are the following: template identification, amino acid sequence alignment, model building and model verification (model quality) [21]. The crystal structure of the transcriptional regulator ATRX from Homo sapiens [22] PDB entry 3qla chain A, was used as a template for homology modeling. The identity between these two proteins was 37% and the structural similarity was 0.42. It is important to mention that the procedures implemented in SWISS-MODEL allow the modeling of sequences which share at least 35% identity with a known three dimensional structure. Although the resulting models do not represent the real 3D structure, it is accurate enough to learn about the general topology and a possible residue arrangement of the ADD protein sequence.

Fly stocks

The wild-type flies used in this study were Oregon R (OreR) or w1118, and fly stocks were maintained at 25°C with standard food. The stocks that carried the atrx alleles were obtained from the Bloomington Stock Center. The stocks that carried dadd1 and Su(var)2-5 alleles were obtained from the Drosophila Genetic Resource Center (DGRC), Kyoto Institute of Technology. The xnp/atrx alleles (called atrx for simplicity) were described by Bassett AR, et al., (2008) [10]. The dadd1 alleles: dadd1NP0793 (w*;P{GawB}NP0793/CyO) and dadd1NP1240 (y*w*;P{GawB}NP1240/CyO, P{UAS lacZ.UW14} UW14), carry a P-element insertion at -225 and -223 bp from the translation initiation codon of the dadd1 gene, respectively. The Su(var)2-52 allele is a missense mutation that has been characterized molecularly by Eissenberg et al., (1992) [23]. This is a single mutation in the open reading frame: a G-A transition in the first nucleotide of codon 26, resulting in the substitution of methionine for valine that affects the chromodomain. The Su(var)2-55 allele is an X-ray induced mutation, in which only the first 10 amino acids of HP1a are translated [23]. The position effect variegation (PEV) reporter line BL1 is an inversion allele of the hsp70-lacZ transgenic reporter, with the reporter gene positioned adjacent to a 3L pericentric heterochromatin mass [24].

Genetic crosses

All stocks were outcrossed with w1118; Sp/CyO; TM6B, Tb/MKRS flies for five generations. Chromosomes with the alleles of interest were followed by segregation with specific balancer chromosomes. To reassure the presence of the atrx alleles in these lines, females carrying alleles atrx1, atrx2 and atrx3 chromosomes were outcrossed with males from parental atrx1 allele and viability was assayed and compared to the previously reported viability [10]. The stocks that carried the atrx alleles were established and balanced with TM6B, Tb for chromosome 3. The stocks that carried the dadd1 alleles were established and balanced with the CyO balancer for chromosome 2 and followed by white complementation. Fly crosses were performed according to standard procedures, three biological replicates were performed. At least 100 flies were examined for each genotype.

Antibodies

The dAtrxL antibody was previously described [25]. All antibodies were generated by New England Peptide (NEP). The pan-dAdd1 antibody was generated using the peptide: QGGEVYCCSTCPYVFCKSC wich recognizes dAdd1-a, b and c isoforms. For the dAdd1-a isoform, the CDLIKALGSPSVLP peptide was used and for the dAdd1-b isoform the CDKQFCQQLVLAM peptide was used. Specificity of these antibodies was assayed by their capability to recognize dAdd1-a or dAdd1-b fused to GST (see pull-down section and antibody specificity assays section for details) by western blot (Fig. S1). The HP1a antibody was obtained from the Developmental Studies Hybridoma Bank DSHB at the University of Iowa (C1A9). The HA antibody was obtained from Roche (Ref.11867423001). Antibodies used for mock immunoprecipitations were purified IgG from mouse GenScript (Cat. A01007); and purified IgG from rabbit Invitrogen (Cat.02-6102).

Antibody specificity assays.

To test the specificity of the pan-dAdd1 antibody, a Western blot was performed with 100 µg of S2R+ protein extracts either using as primary antibody the non-depleted pan-dAdd1 antibody fraction or a supernatant from where the pan-dAdd1 antibodies were depleted (Fig. S1A). Depletion was performed by incubating the pan-dAdd1 antibody in PBS, Tween 0.01%, 5% nonfat milk for 2 hours in the presence of the dAdd1-GST fusion protein blotted onto a nitrocellulose membrane, after the incubation period, supernatant fraction was saved (depleted fraction). The GST-dAdd1 fusion protein harbors the dAdd1 QGGEVYCCSTCPYVFCKSC peptide (aa 122-137) that was used to raise the pan-dAdd1 antibody.

To test the specificity of the dAdd1-a and dAdd1-b antibodies, GST-dAdd1-a or b fusion proteins were expressed in bacteria, blotted onto a nitrocelulose membrane and incubated with either anti-dAdd1-a (Fig. S1B) or anti-dAdd1-b (Fig. S1C) showing that the different antibodies recognize their specific substrate. Detection was performed with the PIERCE quimioluminiscence substrate.

Immunostaining of polytene chromosomes

Polytene chromosomes for immunostaining were obtained from wild type OreR or w1118 lines or from XnpScer\UAS.T:Ivir\HA1/Sgs3-GAL4 (w1118; P{w+mC = Sgs3-GAL4.PD}TP1) larvae. The XnpScer\UAS.T:Ivir\HA1 allele is described in [10], and the Sgs3-GAL4 (w1118; P{w+mC = Sgs3-GAL4.PD}TP1 driver line was used to direct Xnp/dAtrx (called dAtrx for simplicity throughout this manuscript) expression to salivary glands from third instar larvae. Both latter lines were obtained from the Bloomington Drosophila Stock Center.

Immunostaining of polytene chromosomes was performed with slight modifications of the protocol described in [25]. Salivary glands from third instar larvae were fixed in solution I (PBS, 3.7% paraformaldehyde and 1% Triton X-100) and then in solution II (3.7% paraformaldehyde, 50% acetic acid). The chromosomes were spread on poly-L-Lysine coated microscope slides. Anti-HP1a (DSHB) antibody was used at 1∶300, anti-dAtrxL antibody at 1∶100, pan-dAdd1 at 1∶50 and anti-HA (Roche) at 1∶50. Secondary antibodies were Alexa fluor 488 used at 1∶500 and Alexa fluor 568 or 594 used at 1∶100 (Invitrogen). Images were taken on a confocal laser-scanning microscope (Olympus FV1000) at the Laboratoratorio Nacional de Microscopía Avanzada (LNMA, UNAM).

For the double dAdd1/dAtrx immunostaining we followed an epitope tagged version of dAtrxL (dAtrxL-HA) because the dAdd1 and dAtrxL antibodies were raised in the same species and could not be used together for this experiment.

Pull-down assays

We generated fusion proteins of several fragments of dAtrxL and the dAdd1 proteins. All the clones used in this work were nucleotide-sequenced. The Drosophila LD28477, LD24316 and LD37351 cDNA (BDGP Gold collection of Drosophila Genomics Resource Center) were amplified by PCR and cloned in the EcoRI, NotI, SmaI, XhoI or SalI sites of the pGEX-4T1 or pGBKT7 vector (details are available upon request). The dAtrxL and dAdd1 fragments were expressed as GST fusion proteins in a bacterial system. For interaction assays, over-expression of dAtrxL fragments fused to GST in bacteria was induced with 0.4 mM IPTG during 3 h. GST-dAtrxL fragments were purified using glutathione-sepharose (Amersham) according to manufacturer's instructions. The dAdd1 fragments were expressed and labelled with S35 using an in vitro transcription/translation system (TNT-Quick-Coupled Transcription/Translation System from Promega). Pull-down assays were performed according to the manufacturer's instructions.

Cell culture transfection and co-immunoprecipitation

The dAdd1-a cDNA was cloned into EcoRI/NotI sites of pAc5.1/V5-HisA vector (Invitrogen). Drosophila S2R+ cells were maintained in Schneider medium with 10% fetal bovine serum and 100 µg streptomycin/0.25 µg amphotericin. Cells were cotransfected with 10 µg of each construction by the calcium method (Invitrogen). Fourty-eight hours after transfection, the cells were collected and lysed. Lysates were clarified by centrifugation at 15.7 g at 4°C. CoIPs were performed as indicated in [26]. For embryo-stage immunoprecipitations: embryo nuclear extracts were prepared as described [27]. The Co-IP was performed as described in [28]. Antibodies used for mock immunoprecipitations were purified IgG from mouse GenScript (Cat. A01007); and purified IgG from rabbit Invitrogen (Cat.02-6102).

RT–PCR assays

RNA was obtained using the Trizol reagent (Invitrogen®) from embryos (0–3 and 3–21 hour post-fertilization), 1st, 2nd and 3rd instar larvae (L1, L2 and L3), pupae (P), pharate (Ph) and female and male adults Oregon R flies (F and M). 10 µg of total RNA was converted to cDNA using reverse transcriptase enzyme (Invitrogen) and oligo-dT and random primers (Stratagene®). To assess the presence of the transcripts the following oligonucleotides were used for the PCR reaction: dAdd1-a forward (5′CATCTTACGGGCAAAGTGGT-3′); dAdd1-a reverse (5′CAGGCTGGCCAATATCGTGG-3′); dAdd1-b forward (5′GCTTGTCATCGGGCATATCT-3′); dAdd1-b reverse (5′GCTCATAAGCAGCCAGCAGT-3′); dAdd1-c forward (5′ACAGCGGCAGCAACGGAAGC-3′); dAdd1-c reverse (5′GCGGAAGTCCTTGCAGCGGT-3′); rp49 forward (5′TCAAGATGACCATCCGCCCA-3′); rp49 reverse (5′GTTCTCTTGAGAACGCAGGC-3′). rp49 was used as an RT- PCR control.

Chromatin Immunoprecipitation

S2R+ cells or third instar larvae salivary glands from wild type organisms were fixed in 1% formaldehyde. The fixation reaction was stopped by adding glycine (125 mM). Cells or salivary glands were washed and resuspended in lysis buffer, and sonication was performed until the size of chromatin reached between 200 and 800 bp. Pre-clearing, antibody incubations, immunoprecipitation and phenol:chlorophorm extractions were performed as described in [25]. For the ‘mock’ condition, a pre-immune sera against dAtrx or an anti-GFP antibody was used. The following oligonucleotides were used: rover forward (5′-CAACCAAGACCAACCTACCC-3′); rover reverse (5′-GCTCATTTTAGTCTGTCCGC-3′) [29]; for TAS-L, TAS3L_ChIP1 (5′-TGACTGCCTCTCATTCTGTC-3′) and TAS3L_ChIP2 (5′-TATCATCTCGTTCATCCGCC-3′) [30]. qPCRs were performed using the light cycler DNA master SYBR green 1 run in a Lyght cycler 1.5 (ROCHE), and the quantification of %INPUT was performed as in [31].

β-galactosidase quantitative assay

Detection of β-galactosidase in adult flies was performed as in Gu and Elgin (2013) [32]. For quantitative β-galactosidase assays, flies were homogenized in 300 µl of assay buffer (50 mM potassium phosphate, 1 mM MgCl2, pH 7.5), followed by spinning to pellet the debris. An aliquot of 50 µg of protein extract was transferred to CPRG solution (1 mM chlorophenol red b-D-galactopyranoside in assay buffer) and the O.D. at 574 nm was measured at intervals over a 2-hour period. The β-galactosidase activity was calculated as a function of the change in O.D.

Results

The ADD domain of Atrx is highly conserved in Drosophila and other invertebrates

We searched for genes encoding proteins with the ADD domain in the Drosophila genome using the ADD region (aa 169-268) of the hATRX protein with the BLAST (Basic Local Alignment Search Tool) of the NCBI (National Center of Biotechnology Information). Using this approach, we identified the CG8290 annotated gene, which putatively encodes three annotated proteins that have a conserved amino-terminal ADD domain that is 52% homologous and 36% identical to the hATRX ADD domain (Fig. 1A, B). Protein alignment (Fig. 1A) showed that there is a high conservation at the position of the cysteins that coordinate de zinc atoms in the GATA-like and PHD zinc fingers. This conservation is not as high when we align DNMT3L, a methyltransferase that also has an ADD domain (data not shown). The ADD domain in the Drosophila CG8290 encoded proteins is more similar to the one in hATRX than in DNMT3L (data not shown). While this work was in preparation for publication Alekseyenko et al., (2014) named the CG8290 gene dadd1 [11].

In the hATRX protein, there are some amino acids in a "hydrophobic pocket" that mediate the interaction between the H3K9me3 and unmethylated H3K4 (H3K4me0) histone tail combination. In general, it is noteworthy that the whole "hydrophobic pocket" binding site is conserved between the ADD in Drosophila and in hATRX (aa 110-126 in dAdd1) [33]. There are also many conserved amino acids that are not part of the "hydrophobic pocket", such as the histidine 189 and proline 190 of hATRX (which correspond to aa 96 and 97 in dAdd1) that are mutated in ATRX syndrome [22]. We constructed a three-dimensional model of the ADD domain found in the dAdd1 proteins (aa 64 to 164) and compared it to the hATRX ADD domain (Fig. 1B). The human and Drosophila domains mainly overlap in the GATA-like and PHD zinc fingers helices 1, 2 and 3 (h1, h2 and h3, respectively), and the pocket that is important for recognition of H3K9me3 and H3K4me0 is also conserved (Fig. 1C). In the recently published work by Alekseyenko et al., (2014), the capability of this domain to bind to the H3K9me3 tail was assayed, confirming it preferentially binds to this histone modification [11].

We also compared the native structure of the ADD domain of dAdd1 to the ADD domain of DNMT3L (data not shown) and found less overlaping when we superimpose the structures. This led us to propose that during ATRX evolution, the protein may have undergone a fission event. To obtain insight into the evolutionary history of ATRX, we decided to perform a rooted phylogenetic inference analysis using homologous sequences of the ADD and the helicase/ATPase domains from ATRX of higher eukaryotes. We found that the shared common ancestor of higher eukaryotes possessed a protein with both the ADD and helicase/ATPase domains, but in insects, it underwent a fission event by which the two domains were separated, generating two different genes (Fig. 1D, Fig. S2) (perhaps involving gene duplication with subsequent partial degeneration, as has been proposed for cmi and TRR proteins or the monkey king (mkg) gene family in Drosophila) [34], [35].

Another interesting feature is that, in Drosophila, ADD harboring-proteins acquired other domains such as MADF (myb/Sant-like domain in Adf1, shown in blue hexagons), suggesting novel functions for these proteins (Fig. 1D). MADF domains can recognize repetitive sequences on DNA, for instance in Adf1 and Dip3 the MADF domain directs the binding of these transcription factors to specific promoter sequences [36]. All these data led us to propose that the proteins encoded by dadd1 could represent the orthologs of the amino terminal half of the hATRX protein.

The dAdd1 proteins are expressed throughout development and are preferentially nuclear proteins that can bind to different chromatin regions

Through RT-PCR analyses of cDNAs obtained from different developmental stages (see Material and Methods and Fig. 2A), we found three transcripts dAdd1-a, dAdd1-b and dAdd1-c (Fig. 2A, and B top panel). The transcripts (dAdd1-a, dAdd1-b, dAdd1-c) corresponded to the ones derived from alternatively spliced transcripts of CG8290 described in FlyBase [37]. All of these transcripts are deposited maternally into the embryo and are later expressed through all stages of development (Fig. 2B, 0-3 lanes).

thumbnail
Figure 2. The dAdd1 proteins are expressed throughout development.

A. Scheme of the dadd1 mRNAs generated by alternative splicing. The nucleotide sequence of the ADD and MADF domains are represented as gray boxes. Primers used to amplify cDNAs representing mRNAs encoding Add isoforms are indicated by horizontal arrows. B. The dadd1 transcripts were detected throughout all Drosophila stages of development by RT-PCR. Detection of exons from different dadd1 mRNAs in cDNA generated from total RNA isolated from embryos (0–3 and 3–21 hour), 1st, 2nd and 3rd instar larvae (L1, L2 and L3), pupae (P), pharate (Ph) and female and male adults (F and M). rp49 was used as a RT- PCR control. Sequence of the specific primers used to detect and sequence exons from each different dadd1 mRNA are described in Material and Methods section. Molecular weight markers on the left of the panels represent base pairs. C. The dAdd1 (a-c) protein isoforms are mainly nuclear. Detection of dAdd1 (a-c) proteins by Western blotting using the pan-dAdd1 antibody in nuclear (N) and cytoplasmic (C) soluble fractions isolated from S2R+ cells. The largest RNA polymerase II subunit and β-tubulin were used as controls of nuclear and cytoplasmic fractions, respectively. Molecular weight markers on the left of the panels represent kDa. D. dAdd1 proteins co-localize with dAtrx protein in polytene chromosomes of third instar larvae. dAtrxL-HA and dAdd1 were followed with an anti-HA (red, right upper panel) and the pan-Add1 (green, left upper panel) antibodies respectively. The dAdd1 and dAtrxL-HA proteins colocalize in some bands and interbands (left lower panel). They also co-localize in heterochromatic regions such as the chromocenter (see Insets, lower panel right). Note that not all the bands co-localize.

https://doi.org/10.1371/journal.pone.0113182.g002

Interestingly when we amplify the dAdd1-b transcript, another higher molecular weight band appears (Fig. 2B top panel). We sequenced this fragment and it corresponds to a transcript that retains the fifth intron. The presence of this putative transcript varies throughout development being more abundant in larval stages. Bioinformatic analyses indicate that retention of the fifth intron in this transcript would generate a stop codon which will in turn translate into a protein that conserves only one MADF domain. Further analyses are being carried out to assess the presence of this putative protein.

We found that there are three proteins derived from these transcripts that have similar predicted molecular weights ranging from 112 kDa to 130 kDa (Fig. 1C). Interestingly, the dAdd1-b and dAdd1-c proteins have several MADF domains in their carboxy-termini. We designed an antibody that recognizes the three dAdd1 isoforms (called pan-dAdd1, see Materials and Methods), and we tested the antibody's specificity through Western blot and depletion with a GST-dAdd1 fusion protein (Fig. S1A and Materials and Methods section). In the immunoblots we observe two main bands (112 and 130 kDa). Based on the predicted molecular weight of the proteins, the signals of these high-molecular weight bands probably represent the three proteins but cannot be resolved by standard SDS-PAGE electrophoresis.

We found that the dAdd1 proteins are enriched in the nuclear fraction of S2R+ cells (Fig. 2C). Because the proteins are nuclear and have putative chromatin and DNA binding domains, we analyzed their location on chromatin. We made polytene chromosomes preparations from third instar larvae, and using immunostaining with the pan-dAdd1 antibody, we found that the dAdd1 proteins are located in several bands and interbands in the polytene arms (Fig. S3). The signal is also present in heterochromatic regions such as the chromocenter and on the fourth chromosome, and it co-localizes with the dAtrx signal in this region and in other heterochromatic regions (Fig. 2D and Fig. S3).

To determine whether dAdd1 colocalizes with dAtrx, we expressed an epitope tagged-dAtrxL (dAtrxL-HA) [10] in larval salivary glands. This tagged dAtrxL-HA version localizes to the same sites as wild type dAtrxL [25] and does not alter dAdd1 distribution (compare Fig. 2D and Fig. S3). We found that dAdd1 colocalizes with dAtrxL-HA in the chromocenter and in other heterochromatic regions (Fig. 2D insets).

These results opened the possibility that dAdd1 proteins with an ADD domain could interact with the ADD-less Drosophila dAtrx and cooperate in some cellular functions in the organism.

The dAdd1 proteins interact directly with the dAtrxL protein

To test the proposed interaction hypothesis, we performed co-immunoprecipitation experiments using different antibodies generated against the dAdd1 proteins and dAtrxL (see Materials and Methods). As a first approach we coimmunoprecipitated dAdd1 with the pan-dAdd1 antibody and tested whether we could detect dAtrxL. In addition to detecting dAdd1 (Fig. 3A, top panel, pan-dAdd1 blot) we found dAtrxL in the immunoprecipitate (Fig. 3A top panel, dAtrxL blot, IP lane), the specific bands obtained in the IP are marked with a white arrow. We also performed the reciprocal immunoprecipitation with the dAtrxL antibody, and we show that at least one of the dAdd1 proteins, approximately 130 kDa, co-immunoprecipitates with dAtrxL (Fig. 3A bottom panel, IP lane) the specific band is marked with a black arrow. The signal in both cases is specific because it cannot be detected when we used an irrelevant antibody to perform the immunoprecipitation assay (mock lane in all the blots).

thumbnail
Figure 3. The dAtrx and dAdd1 proteins physically interact.

A. Co-immunoprecipitation (CoIP) assays of dAdd1 and dAtrxL. IP performed with pan-dAdd1 antibody (top panels) and the reciprocal CoIP assay with anti-dAtrxL antibody (bottom panels). Embryo nuclear extract (NE); S2R+ cells total extract (TE); pre-clearing (PC); unbound protein fraction (UB); Mock is the IP performed with an unrelated antibody. The presence of dAdd1 and dAtrxL in the precipitated proteins was determined by Western blotting using the specific antibody. 10% of the nuclear extract (corresponding to approximately 50 µg of protein) used for the IP was loaded as the INPUT fraction. 20% of the total extract (corresponding to 150 µg of protein) was loaded as the INPUT fraction. Molecular weight markers on the left of the panels represent kDa. B. CoIP assay of dAdd1-a and dAdd1-b with dAtrxL. S2R+ cells total extract (TE); pre-clearing (PC); unbound protein fraction (UB); IP was performed with anti-dAtrxL. Mock is the IP performed with an unrelated antibody. The presence of dAdd1-a and dAdd1-b in the precipitated proteins was determined using the specific antibody. 20% of the total extract (corresponding to 150 µg of protein) was loaded as the INPUT fraction. The 250 kDa band observed in the INPUT lane is an unspecific band. Molecular weight markers on the left of the panels represent kDa. C. Representation of the fragments assayed by Pull-down. The important fragments for the interaction between dAtrx and dAdd are shown in dashed boxes. D. Pull-down assay. The first lane of all panels (except the coomassie panels) shows the 10% of the total amount of transcription/translation labeled protein (dAdd1 fragments) used for each experiment (input). The rest of the lanes are the experimental interaction (GST-dAtrx fragments and dAdd1 fragments) and the control interaction (GST and dAdd fragments) for each analyzed polypeptide. Coomasie staining of the loaded GST-fused proteins is shown in the bottom panels.

https://doi.org/10.1371/journal.pone.0113182.g003

To further characterize which of the dAdd1 proteins are in the observed immunoprecipitated band, we performed a coimmunoprecipitation assay using S2R+ total extracts with antibodies generated against dAdd1-a or dAdd1-b. We found that both, dAdd1-a and dAdd1-b coimmunoprecipitate with dAtrxL (Fig. 3B, top and bottom panels), the white arrows indicate the specific bands corresponding to the co-immunoprecipitated proteins. In the dAdd1-a blot (Fig. 3B, top panel) there is a high molecular weight band, near the 250 kDa weight marker, this band is unspecific because it also appears in the immunoprecipitation performed with a purified rabbit IgG antibody (Fig. 3B, mock lane). The fact that both proteins, dAdd1-a and dAdd1-b, can immunoprecipitate with dAtrxL indicates that the interaction domain is most likely conserved in both proteins or that a third protein is mediating the interaction.

To map the dAdd1 protein domain(s) important for the interaction with dAtrxL, we performed a series of pull-down assays. We generated several fragments of dAtrxL and dAdd1 proteins and expressed them as GST-fusion proteins in a bacterial system or expressed and labeled them with S35 using an in vitro transcription/translation system (see Materials and Methods, Fig. 3C). We tested four different fragments of the dAdd1-a protein. Two of these fragments are present in the three isoforms (dAdd1 1-225, dAdd1 1-625). When we assayed the dAdd1 1-225 fragment with three different dAtrxL fragments, we observed that it binds to all of the tested fragments of dAtrxL, but this fragment also binds to the negative control GST, although with less affinity (Fig. 3D top panel on the right). In contrast, the larger fragment that contains aminoacids 226-625 (fragment dAdd1 1-625) binds specifically to the amino-terminal fragment of dAtrxL (dAtrx 1-221) (Fig. 3D top left panel).

We also tested the dAdd1-a (355-1199) and dAdd1-a (620-1199) fragments. The first fragment, which contains amino acids 355-625 binds to two dAtrxL fragments (1-221 and 368-1103), corresponding to the amino-terminal and SNF2 fragments, whereas the second dAdd1-a fragment (620-1199), which lacks amino acids 355-625, fails to interact with any of the tested dAtrxL fragments (compare second and third panels from top to bottom on the right of Fig. 3D).

In these analyses, we also found that a fragment that includes the MADF domains (dAdd1-b 538-1125) interacts with the amino terminal and SNF2-containing fragments of the dAtrxL protein (Fig. 3D from top to bottom: fourth panel on the right) but this fragment also binds to the negative control GST.

From these data, we conclude that the dAdd1 domain that interacts with dAtrxL is conserved in all dAdd1 isoforms. Thus, all dAdd1 isoforms can directly interact with the amino-terminal domain of dAtrxL (a.a. 1-221) through their amino-terminal domain (a.a. 1-625). Interestingly, this dAtrxL fragment is only conserved in the long isoform, and it does not overlap with the HP1a interaction domain [22]. In contrast, it does overlap with the DREF interaction domain, which was reported by our group [25]. Based on these results, we conclude all dAdd1 isoforms can interact directly with the dAtrxL protein.

atrx and dadd1 interact genetically

At this point we have demonstrated that dAdd1 and dAtrx physically interact; and therefore, we wanted to know whether they genetically interact in the fly. For this purpose, we made transheterozygous atrx and dadd1 flies. For these analyses, we used three atrx alleles [20]. atrx1 is a deficiency that uncovers dAtrx as well as three more adjacent genes; atrx2 is an hypomorphic allele that affects both the long and short dAtrx isoforms and atrx3 is an hypomorphic allele that affects only the long isoform. The survival of heteroallelic atrx flies is affected, as previously reported, and they do not present any other obvious phenotype (Table 1, [10]).

We also used two P-element insertion lines, dadd1NP1240 and dadd1NP0793. These insertions lie at 83 or 81 bp, respectively, from the dadd1 transcriptional start site (Fig. 4A). To determine whether the insertions affected the levels of the dAdd1 proteins, we extracted total proteins from adult flies of the different genotypes and analyzed the presence of the dAdd1 proteins through Western blotting using the pan-dAdd1 antibody (Fig. 4B). We found that the levels of the dAdd1 proteins are diminished in homozygous dadd1NP1240/dadd1NP1240 individuals, with respect to the heterozygous dadd1NP1240/+ or +/+ flies (Fig. 4B). We also analyzed the dadd1 mRNA levels of dadd1NP1240/+ or dadd1NP0793/+ individuals through semi-quantitative RT-PCR and found that the dadd1 mRNA levels are diminished in the mutant individuals compared to the dadd1 wild type flies (Fig. S4). These results indicate that these alleles are hypomorphic. Similar results were obtained with the dadd1NP1240/dadd1NP0793 heteroallelic flies (data not shown) and thus we used for the rest of our tests the dadd1NP1240 allele. 44% and 90% of the dadd1NP1240/dadd1NP1240 and dadd1NP1240/dadd1NP0793 flies reach adulthood respectively, showing that the higher lethality present in dadd1NP1240/dadd1NP1240 individuals (56%) compared to the one presented by the dadd1NP1240/dadd1NP0793 flies (only 10%) may be caused by other lethals present in the dadd1NP1240 chromosome (Table 1). The heteroallelic dadd1NP1240/dadd1NP0793 individuals are fertile and do not present any other obvious phenotype.

thumbnail
Figure 4. Genetic interaction between dadd1 and atrx.

A. Scheme of dadd1 showing the position of the EP elements insertions (gray triangle) in dadd1NP1240 and dadd1NP0793 alleles, introns are represented by white boxes while exons are represented by gray boxes. The start codon at position 308 pb is also shown. B. dadd1NP1240 is an hypomophic dadd1 allele. Western blotting assay probed with the pan-dAdd1 antibody. Each lane contains 150 µg of proteins from adult flies of the indicated genotype. β-tubulin was used as a loading control. Note the low levels of the dAdd1 proteins in homozygous individuals with respect to +/+ individuals. This demonstrates the allele is hypomorphic. The 250 kDa band observed in the wild type lane is an unspecific band. Molecular weight markers on the left of the panels represent kDa. C. Some dadd1/atrx individuals present melanotic masses. Melanotic masses that appear during larval (upper panel) and adult stages (lower panel) of mutant dadd1/atrx individuals are shown. Wild type individuals (w1118 for larvae and OreR for adult) are shown to the left for comparison.

https://doi.org/10.1371/journal.pone.0113182.g004

Survival of the transheterozygous dadd1/atrx flies is mildly affected, and a small percentage of these flies present melanotic masses, both in third instar larvae and adult individuals. The percentage of melanotic masses is higher in flies carrying two different alleles of atrx in combination with one allele of dadd1 (Fig. 4, Table 1). We obtained similar results for all the combinations of atrx and dadd1 alleles (Fig. 4, Table 1,). The masses also appear during larval stages, and these individuals fail to advance further in development (Fig. 4C). Similarly, adult flies presenting melanotic tumors die within the first 10 days post-eclosion (data not shown). These data provide evidence that, besides the physical interaction we showed in the previous sections, there is a genetic interaction between dAdd1 and Atrx functions that is essential for fly development.

The dAdd1 proteins co-localize with HP1a in heterochromatic regions and cooperate with dAtrx in the maintenance of pericentric heterochromatin

As dAdd1 proteins interact with dAtrxL, we thought that they could be involved in heterochromatin maintenance; therefore, we analyzed their interaction with HP1a.

As mentioned before, the HP1α interaction domain in the dAtrxL protein (245aa-CxVxL-249aa, [10]) does not overlap with the interaction domains of the dAdd1 proteins. It is possible that dAdd1-a, b or c, dAtrxL and HP1a can co-exist in the same protein complex. If this is true, then we should be able to coimmunoprecipitate HP1α with the dAdd1 proteins. We expressed a V5 epitope-tagged dAdd1 version in S2R+ cells and detected the presence of HP1a in an immunoprecipitation (IP) assay of the V5-dAdd1 protein with the V5 antibody (Fig. 5A, lower panel, white asterisk). Although the signal is faint it is specific because it cannot be detected in the mock control IP (Fig. 5A). This finding indicates that dAdd1-a is able to co-immunoprecipitate with HP1a.

thumbnail
Figure 5. The dAdd1 proteins co-localize in heterochromatic regions with HP1a and cooperate with dAtrx in the maintenance of pericentric heterochromatin.

A. CoIP assay of dAdd1-a and HP1a. Total extract (TE) from transiently transfected S2R+ cells with dAdd1-a-V5, P1 and P2 are pre-clearing 1 and 2 (see Material and Methods); unbound protein fraction (UB); IP performed with anti-V5 (upper panel); Mock is the IP performed with purified mouse IgG. The presence of dAdd1-a-V5 was determined using the anti-V5 antibody. The presence of HP1a was determined using the anti-HP1a (CIA9) antibody (lower panel). Molecular weight markers on the left of the panels represent kDa. B. Polytene chromosomes preparations from wild-type OreR third instar larvae were simultaneously stained with anti-HP1a (red signal) and pan-dAdd1 (a-c) antibody (green signal). Merge bands containing both factors can be visualized mainly in the chromocenter (yellow signal). C. dadd1 is a suppressor of position effect variegation. Quantitative β-galactosidase assay. The graphs show that dadd1 alleles can suppress position effect variegation measured by the enzymatic activity of Lac-Z. WT =  wild type flies with CyO and TM3 chromosome balancers; C =  control flies derived from F1 progeny of paternally delivered dadd1 alleles; M =  Maternal depletion, F1 progeny flies of maternally delivered dadd1 alleles; Z = zygotic depletion, flies derived from crossing males carrying the dadd1 alleles to females carrying the BL1 allele; M+Z =  maternal and zygotic depletion, flies derived from crossing females carrying the dadd1 alleles to males carrying the BL1 reporter allele. The bottom graph shows that transheterozygous dadd1, atrx flies have an additive effect in the suppression of position effect variegation.

https://doi.org/10.1371/journal.pone.0113182.g005

We also show that dAdd1 and HP1a co-localize in heterochromatin regions and in the chromocenter of wild type polytene chromosomes (Fig. 5B yellow arrow). The dAdd1 signal is also observed in regions where no HP1a signal is detected (Fig. 5B green arrow).

The coimmunoprecipitation and immunolocalization analyses suggest that the dAdd1 proteins could be located in different regions of the chromatin. To analyze the binding of these factors at several heterochromatic and euchromatic regions in more detail, we performed chromatin immunoprecipitation (ChIP) assays using different antibodies against HP1a, dAtrxL and the dAdd1 proteins with chromatin from third instar larvae salivary glands and S2R+ cells (see Material and Methods).

HP1α is not enriched in the TAS (Telomeric Associated Sequences) regions [38]. On the other hand, Antão et al., (2012) [30] found through a proteomic analysis of chromatin segments (PICh) that dAtrx and of the product of CG8290 are enriched in the telomeric associated sequences (TAS-L) but they did not find HP1a. Thus, we looked for the presence of HP1α, dAtrxL and the dAdd1 proteins in the telomeric region (TAS-L). Our ChIP experiments show, similar to Antão et al., (2012) [30], that dAtrxL and the proteins, but not HP1α, are enriched in the TAS-L region (Fig. S5). We also found that dAtrxL and the dAdd1 proteins colocalize in a region that can potentially form a G quadruplex structure (G4) (Fig. S5).

In search for heterochromatic regions where the proteins did not co-localize we tested the rover retrotransposon, which we know is a constitutive heterochromatic region enriched in H3K9me3 mark [29]. Here we found dAtrxL is present but the dAdd1 proteins are not enriched in this region. This result indicates that dAdd1 and dAtrxL do not colocalize at all the heterochromatic regions in the genome (Fig. S5). Next, we looked for an euchromatic region, we decided to use the sgs8 promoter which is actively transcribing in third instar salivary glands [29]. We observed that in this euchromatic region there is some enrichment of dAdd1 proteins whereas dAtrxL is not enriched as expected (Fig. S5). This finding is supported by the fact that in the immunolocalization experiments shown above, we also observe dAdd1 in euchromatic regions in the chromosome arms (Fig. 2D and Fig. S3).

The results obtained in the ChIP analyses could represent the capability of the different dAdd1 proteins to bind different regions of the chromatin. The fact that dAtrxL and the dAdd1 proteins are not always together, as shown in the ChIP and the immunolocalization analyses, provide evidence that the dAdd1 proteins have functions independent of that with dAtrxL. This result also shows that dAtrxL can bind to chromatin regions independent of the dAdd1 proteins, most likely through interactions with other factors such as DREF [23].

The colocalization of HP1a, dAtrxL and dAdd1 in the chromocenter raises the possibility that the dAdd1 proteins are involved in the maintenance of heterochromatin. In a model for the maintenance of pericentric heterochromatin proposed by Eustermann S, et al., (2011) [4], the ADD domain of the hATRX protein, reinforced by the interaction with HP1a, recognizes pericentric heterochromatin. Once hATRX is recruited its helicase/ATPAse domain directs deposition of the histone variant H3.3 [4].

It has not been determined conclusively whether Drosophila dAtrx is involved in the deposition of the histone variant H3.3, but both HP1a and dAtrxL localize at the chromocenter of polytene chromosomes and act as suppressors of position effect variegation for the wm4 allele, an inversion that lies near pericentric heterochromatin [10].

Following this hypothesis we tested the dadd1 capability to suppress position effect variegation. We used the BL1 line, which carries a LacZ reporter construct that lies near the centromere [24], [32]. In this assay, we can determine whether dadd1 is involved in the establishment, crossing virgin females carrying the dadd1 alleles to males carrying the reporter BL1 (maternal effect), or maintenance, crossing males carrying the dadd1 alleles to females carrying the reporter BL1 (zygotic effect), of heterochromatin [32] (see Material and Methods).

The results show that dadd1 does not have a relevant role in the establishment of pericentric heterochromatin (compare M (maternal) bar and C (control) bar in Fig. 5C top graph). When we analyzed the zygotic effect, virgin BL1 females crossed to dadd1 male carrying alleles, the activity of LacZ was consistently higher in dadd1 mutants (compare Z (zygotic) bar to C (control) bar in Fig. 5C top graph), indicating that the dadd1 alleles can suppress position effect variegation and that dadd1 has a role in the maintenance of heterochromatin rather than in the establishment of heterochromatin.

Next, we tested whether a combination of dadd1 and atrx alleles enhances the suppression of position effect variegation in a cooperative manner. In fact, this is what we observed. Our results show that transheterozygous dadd1/atrx flies suppress position effect variegation more efficiently than heteroallelic flies carrying either the dadd1 or atrx alleles (compare third and fourth bars in Fig. 5C bottom graph).

Overall, these results indicate that dAdd1 proteins are capable of interacting with proteins, such as HP1a, that are involved in the establishment of heterochromatin and that they cooperate with dAtrxL in the maintenance of heterochromatin.

Su(var)205 interacts genetically with dadd1, and atrx

We showed that dadd1/atrx individuals present melanotic masses. The Su(var)2052 is a loss-of-function HP1a allele because it carries a mutation in aminoacid 26 within the chromodomain. We tested whether the loss of function of HP1a in the presence of dadd1 and/or atrx mutations affects the presence of the melanotic masses found in dadd1/atrx individuals.

Transheterozygous Su(var)2052 flies carrying the dadd1, or atrx1 or atrx2 alleles do not present melanotic masses, whereas Su(var)2052 flies carrying atrx3 (which affects only the long isoform of dAtrx) present melanotic masses in a small percentage (Table 2). The penetrance of melanotic masses in daddNP1240/+; atrx3/+ (4%) or Su(var)2052/+; atrx3/+ (2%) individuals increases to 14% in Su(var)2052/daddNP1240;atrx3/+ individuals where mutations in the three genes are together (Table 2). These data show that the presence of melanotic masses in transheterozygous dadd1, atrx and Su(var)205 flies most likely involves misregulation in the maintenance of heterochromatin.

Discussion

In vertebrates, hATRX is a protein with an ADD domain and an SNF2-helicase/ATPase motif. Through evolution, the SNF2 domain of hATRX has been highly conserved, but in invertebrates, the ADD domain is lost (Fig. 1D, [39]). In this article, we describe the characterization of the Drosophila dAdd1 proteins as orthologs to the amino-terminal region of Atrx. The dadd1 gene expresses the dAdd1 proteins, which have an ADD domain (Fig. 1C), throughout development (Fig. 2B). The 3D structure of the dAdd proteins ADD domain (Fig. 1B) is more similar to the one found in hATRX than to the ADD domains found in other proteins. Drosophila dAdd1 and the ADD-less dAtrx interact physically (Fig. 3A, 3D) and genetically (Tables 1 and 2), and they interact with HP1a in heterochromatic regions (Fig. 5B).

The human hATRX gene encodes an SNF2 helicase/ATPase protein that has many different functions. It is part of a complex that includes the histone variant H3.3 chaperone DAXX, and it is involved in the deposition of this histone variant in pericentric and telomeric heterochromatin [7], [40], [41]. The hATRX protein can also bind to regions in the genome that can potentially form G-quadruplex structures and it has been proposed that this binding alleviates the quadruplex conformation and allows the deposition of H3.3 [9], [42]. Its participation in the deposition of histone variant H3.3 requires the helicase/ATPase activity, although it has not been determined whether this activity is also required for the binding and recognition of the G-quadruplex structures of the DNA. In addition to the SNF2 helicase/ATPase domain, hATRX also has an amino-terminal domain composed of two zinc fingers, a GATA-like finger and a PHD finger called the ADD domain. This domain is able to recognize the H3K9me3 histone mark in combination with unmodified H3K4 through a "hydrophobic pocket", which is an unusual feature for a PHD zinc finger [33]. The interaction between hATRX and histone H3 chromatin marks has been shown to maintain hATRX binding to pericentric heterochromatin. A model has been proposed in which hATRX is directed to pericentric heterochromatin through the ADD domain, and this interaction is reinforced by the interaction between hATRX and HP1a. In this model, the SNF2 helicase/ATPase domain then directs histone H3.3 deposition or performs other required ATP-dependent functions [4].

dadd1 encodes three polypeptides that contain an ADD domain that is highly similar to the hATRX ADD domain (36% identity and 52% of similarity, [11] and this work). The comparison of the 3D-structure analyses of the ADD of Drosophila dAdd1 and hATRX reveal that these domains overlap and that the histone recognition pocket is conserved.

Our phylogenetic inference analysis using the ADD domain of hATRX indicates that the ATRX gene underwent a fission event. This fission has been described in at least another chromatin-binding-protein encoding gene, cara mitad (cmi), which is the homolog of the amino-terminal portions of mammalian MLL2 and MLL3 [32].

It is intriguing that there are three spliced isoforms encoded by the dadd1 gene. The dAdd1-a isoform contains the ADD domain and no other putative domains, while the other two spliced isoforms have additional domains called MADF (myb/SANT-like domain in Adf1). The MADF domains are a subfamily of the SANT domains. They are DNA-binding domains found in all Drosophila species. In the case of Adf1, the first Drosophila factor identified to have an homology to the Myb oncoprotein [43], the MADF domain recognizes and binds certain sequences in repetitive regions [44], although there are some examples in which the MADF domains can also bind proteins, and apparently the specificity lies in the domain's isoelectric point [45].

The predicted isoelectrical point of the three MADF domains in the dAdd1 proteins is basic, which could indicate that they can bind to DNA.

The acquisition of new domains indicates that these proteins most likely have diverged functions. One indication of this could be the pattern of the immunolocalization of dAdd1 in polytene chromosomes, where not all the signals derived from the pan-dAdd1 antibody co-localize with dAtrxL. This indicates that the dAdd1 isoforms could also have roles independent of their interaction with dAtrxL.

The survival of transheterozygous individuals carrying the dadd1/atrx alleles is compromised, and some of the larvae and adult individuals show melanotic masses. We found that these proteins also colocalize with HP1a and that dAdd1-a immunoprecipitates with HP1a. Diminished HP1a levels in dadd1/atrx flies enhance the incidence of melanotic masses. Our results are supported by the recent report of Alekseyenko et al., (2014), in which it is demonstrated that dAdd1 physically interacts with HP1a and that dAdd1 is a suppressor of variegation [11].

The generation of melanotic masses involves problems in the differentiation of hematopoietic cells in the flies. The differentiation of hematopoietic cells takes place in the early embryonic head mesoderm and in the lymph gland of late larvae, and three different types of cells are derived from the prohemocytes; the plasmatocytes, crystal cells and lamellocytes. Several different pathways regulate the differentiation of these cells [46]. One of the pathways involves a balance in the expression of the Pnr-α and Pnr-β proteins, which is controlled by the JAK/STAT pathway [48]. Jak hyperactivation results in the proliferation of hemocytes, lamellocyte differentiation and the generation of melanized pseudotumors. Pseudotumors or melanotic masses are formed by crystal cells and lamellocytes [47]. On the other hand, Stat is a positive regulator of plasmatocyte differentiation, and one of the downstream factors regulated by Stat is the GATA factor Pannier [48]. Importantly, our group had already identified pannier (pnr) as a gene regulated by dAtrxL and another transcriptional factor DREF [25]. In that report, we concluded that dAtrxL is recruited to the pnr gene promoter through DREF, and that it acts as a co-repressor of pnr gene expression.

Furthermore, the data obtained from the genetic interactions between the dadd1, atrx and Su(var)2052 alleles indicates that melanotic masses are derived from problems in the proteins involved in heterochromatin maintenance, such as dAtrxL and HP1a. The position effect variegation assay indicated that the dAdd1 proteins are involved in the maintenance of heterochromatin. Thus, an attractive hypothesis is that dAdd1, dAtrxL and HP1a cooperate and promote heterochromatinization at the promoters of the genes involved in the JAK/STAT pathway, including pnr. Lack of these proteins could lead to misregulation of the effectors of the pathway, giving rise to the melanotic masses. Experiments are being carried out to test this hypothesis.

In the human ATRX syndrome, the majority of the mutations identified so far affect the ADD or the SNF2 helicase/ATPAse domains. The fact that the domains are separated in flies provides a new important tool to study the individual roles these domains have in the development of the organism.

Supporting Information

Figure S1.

dAdd1 antibodies recognize specifically the dAdd1 proteins. a) Western blot using the pan-dAdd1 antibody. The dAdd1 isoform signals (lane 1) observed with the pan-dAdd1 antibody are no longer observed in lane 2. This demonstrates that the GST-dAdd1 fusion protein which harbors the peptide used to raise the pan-dAdd1 antibody is able to deplete them from this fraction. The dAdd1 signals are no longer observed (lane 2) showing that the pand-dAdd1 specifically recognizes the dAdd1 isoforms (see also Materials and Methods). b) Anti-dAdd1-a antibody recognizes the dAdd1-a protein. Indicated GST fusion proteins were loaded and blotted onto a nitrocelulose membrane. The Western blot was performed with an anti-dAdd1-a antibody (top panel) or an anti-GST antibody (bottom panel). The dAdd1 specific signal is observed only where the GST-dAdda-1 fusion protein harboring the peptide used to raise the antibody was loaded (GST-dAdd1 aminoacids 620-1199) (left lane, top panel). The antibody does not recognize a fusion protein that lacks this peptide (right lane, top panel). The GST antibody recognizes the aforementioned two GST-dAdd1 fusion proteins (bottom panel). Extra bands (asterisks, right lane, bottom panel) may be fusion protein degradation. c) Specificity test for the dAdd1-b antibody. Indicated GST-fusion proteins were induced in E. coli. Induced extracts were loaded and blotted onto nitrocellulose membranes. The Western blot (upper panel) was performed using the anti-dAdd1-b antibody. A specific signal is observed (lane 3, black arrow in the blot and in the Ponceau staining) which corresponds to the induced GST-dAdd1-b fragment (aa 530-1125) and not GST or GST-dAdd1-a (aa 620-1199) (lanes 1 and 2 respectively). Faint bands in the first two lanes are unspecific signals.

https://doi.org/10.1371/journal.pone.0113182.s001

(TIF)

Figure S2.

The atrx gene suffered a fission event in the Insecta class. Maximum Likelihood Phylogenetic Analysis of Helicase/ATPase domain and the corresponding Protein Domain Architecture information of its containing proteins. The numbers shown represent bootstrap values. It can be seen that the common ancestor to plants and animals had a protein with both, the ADD and the helicase/ATPase domains, but insects show the homologous domains in different proteins. Since it is more likely that only one fusion event, instead of two independent acquisitions of the same domain, occurred during the evolution, the most parsimonious explanation is to consider a model in which a gene fission event occurred within the Insecta class. For the parameters used, see Material and Methods section.

https://doi.org/10.1371/journal.pone.0113182.s002

(DOCX)

Figure S3.

The dAdd1 proteins localize at many chromatin regions in polytene chromosomes. Wild type polytene chromosome staining was performed with the pan-dAdd1 antibody (red). The dAdd1 proteins localize in heterochromatic regions such as the chromocenter and the fourth chromosome (inset, white arrow). Staining along the chromosomes arms and in euchromatic regions is also observed.

https://doi.org/10.1371/journal.pone.0113182.s003

(TIF)

Figure S4.

Alleles dadd1NP1240 and dadd1NP0793 are hypomorphs. Semiquantitative RT-PCR from wild type and mutant dadd1 flies. dadd1 mRNA level is lower in the dadd1NP1240 (upper panel) and dadd1NP0793 (lower panel) heterozygous flies than in the dadd1 wild type flies (w1118). In homozygous dadd1NP1240/dadd1NP1240 (upper panel), dadd1 is even lower that in the heterozygous condition. rp49 transcript levels remained unchanged in the mutant alleles.

https://doi.org/10.1371/journal.pone.0113182.s004

(DOCX)

Figure S5.

The dAdd1 proteins co-localize in vivo with dAtrxL and HP1a in some chromatin regions. ChIP assay using total extracts from third instar salivary glands (SG) prepared from wild-type larvae and S2R+ cells. Graphs represent the percentage of input precipitated using the different antibodies for the same regions. Note that in the rover region only dAtrxL is enriched. Three independent biological replicates were performed.

https://doi.org/10.1371/journal.pone.0113182.s005

(TIF)

Acknowledgments

We would like to thank Dr. Sarah Elgin for kindly providing the BL1 reporter fly stock. Stocks obtained from the Bloomington Drosophila Stock Center (NIH P40OD018537) were used in this study. We also thank Carmen Muñoz and the Laboratorio Nacional de Microscopía Avanzada (LNMA) of IBt, UNAM for technical assistance.

Author Contributions

Conceived and designed the experiments: MZ VVG. Performed the experiments: BLF SMN BHR AC DM VVG. Analyzed the data: ER MZ MV VVG. Contributed reagents/materials/analysis tools: AC ER. Wrote the paper: MZ MV VVG.

References

  1. 1. Gibbons RJ (2012) α-Thalassemia, Mental Retardation, and Myelodysplastic Syndrome. Cold Spring Harb Perspect Med 2: pii: a011759.
  2. 2. Garrick D, Samara V, McDowell TL, Smith AJ, Dobbie L, et al. (2004) A conserved truncated isoform of the ATR-X syndrome protein lacking the SWI/SNF-homology domain. Gene 326:23–34.
  3. 3. Dhayalan A, Tamas R, Bock I, Tattermusch A, Dimitrova E, et al. (2011) The ATRX-ADD domain binds to H3 tail peptides and reads the combined methylation state of K4 and K9. Hum Mol Genet 20:2195–2203.
  4. 4. Eustermann S, Yang JC, Law MJ, Amos R, Chapman LM, et al. (2011) Combinatorial readout of histone H3 modifications specifies localization of ATRX to heterochromatin. Nat Struct Mol Biol 18:777–782.
  5. 5. Mitson M, Kelley LA, Sternberg MJ, Higgs DR, Gibbons RJ (2011) Functional significance of mutations in the Snf2 domain of ATRX. Hum Mol Genet 20:2603–2610.
  6. 6. Emelyanov AD, Konev YA, Vershilova E, Fyodorov DV (2010) Protein complex of Drosophila ATRX/XNP and HP1a is required for the formation of pericentric beta-heterochromatin in vivo. J Biol Chem 285:15027–15037.
  7. 7. Gibbons RJ, Wada T, Fisher CA, Malik N, Mitson MJ, et al. (2008) Mutations in the chromatin-associated protein ATRX. Hum Mutat 29:796–802.
  8. 8. Goldberg AD, Banaszynski LA, Noh KM, Lewis PW, Elsaesser SJ, et al. (2010) Distinct factors control histone variant H3.3 localization at specific genomic regions. Cell 140:678–691.
  9. 9. Law MJ, Lower KM, Voon HP, Hughes JR, Garrick D, et al. (2010) ATR-X syndrome protein targets tandem repeats and influences allele-specific expression in a size dependent manner. Cell 143:367–378.
  10. 10. Bassett AR, Cooper SE, Ragab A, Travers AA (2008) The chromatin remodelling factor dATRX is involved in heterochromatin formation. PLoS ONE 3(5):e2099.
  11. 11. Alekseyenko AA, Gorchakov AA, Zee BM, Fuchs SM, Kharchenko PV, et al. (2014) Heterochromatin-associated interactions of Drosophila HP1a with dADD1, HIPP1, and repetitive RNAs. Genes Dev 28:1445–60.
  12. 12. Ren J, Longping W, Gao X, Jin C, Xue Y, et al. (2009) DOG 1.0: Illustrator of protein domain structures. Cell Res 19:271–273.
  13. 13. Rambaldi D, Ciccarelli FD (2009) FancyGene: dynamic visualization of gene structures and protein domain architectures on genomic loci. Bioinformatics 25:2281–2282.
  14. 14. Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan PA, et al. (2007) Clustal W and Clustal X version 2.0. Bioinformatics 23:2947–2948.
  15. 15. Abascal F, Zardoya R, Posada D (2005) ProtTest: Selection of best-fit models of protein evolution. Bioinformatics 21:2104–2105.
  16. 16. Guindon S, Gascuel O (2003) A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Syst Biol 52:696–704.
  17. 17. Le SQ, Gascuel O (2001) An Improved General Amino Acid Replacement Matrix. Mol Biol Evol 25:1307–1320.
  18. 18. Letunic I, Bork P (2011) Interactive Tree Of Life v2: online annotation and display of phylogenetic trees made easy. Nucleic Acids Res 39 : (Web Server issue): W475–8.
  19. 19. De Castro E, Sigrist CJA, Gattiker A, Bulliard V, Langendijk-Genevaux PS, et al.. (2006) ScanProsite: detection of PROSITE signature matches and ProRule-associated functional and structural residues in proteins. Nucleic Acids Res 34 (Web server issue):W362–W365.
  20. 20. Arnold K, Bordoli L, Kopp J, Schwede T (2006) The SWISS-MODEL workspace: a web-based environment for protein structure homology modelling. Bioinformatics 22:195–201.
  21. 21. Biasini M, Bienert S, Waterhouse A, Arnold K, Studer G, et al.. (2014) SWISS-MODEL: modelling protein tertiary and quaternary structure using evolutionary information. Nucleic Acids Res 42 :(Web server issue): W252–8.
  22. 22. Iwase S, Xiang B, Ghosh S, Ren T, Lewis PW, et al. (2011) ATRX ADD domain links an atypical histone methylation recognition mechanism to human mental-retardation syndrome. Nat Struct Mol Biol 18:769–76.
  23. 23. Eissenberg JC, Morris G, Reuter G, Hartnett T (1992) The heterochromatin associated protein HP1 is an essential protein in Drosophila with dosage dependent effects on position effect variegation. Genetics 131:345–352.
  24. 24. Lu BY, Bishop CP, Eissenberg JC (1996) Developmental timing and tissue specificity of heterochromatin-mediated silencing. EMBO J 15:1323–1332.
  25. 25. Valadez-Graham V, Yoshioka Y, Velázquez O, Kawamori A, Vázquez M, et al. (2012) XNP/dATRX interacts with DREF in the chromatin to regulate gene expression. Nucleic Acids Res 40:1460–74.
  26. 26. Herrera-Cruz M, Cruz G, Valadez-Graham V, Fregoso-Lomas M, Villicaña C, et al. (2012) Physical and functional interactions between Drosophila homologue of Swc6/p18 Hamlet subunit of the SWR1/SRCAP chromatin-remodeling complex with the DNA repair/transcription factor TFIIH. J Biol Chem 287:33567–33580.
  27. 27. Kamakaka RT, Kadonaga JT (1994) The soluble nuclear fraction, a highly efficient transcription extract from Drosophila embryos. Methods Cell Biol 44:225–235.
  28. 28. Leclerc V, Tassan JP, O'Farrell PH, Nigg EA, Leopold P (1996) Drosophila Cdk8, a kinase partner of cyclin C that interacts with large subunit of RNA polymerase II. Mol Biol Cell 7:505–513.
  29. 29. Palomera-Sanchez Z, Bucio-Mendez A, Valadez-Graham V, Reynaud E, Zurita M (2010) Drosophila p53 is required to increase the levels of the dKDM4B demethylase after UV induced DNA damage to demethylate histone H3-lysine 9. J Biol Chem 285:31370–31379.
  30. 30. Antão JM, Mason JM, Déjardin J, Kingston RE (2012) Protein landscape at Drosophila melanogaster telomere-associated sequence repeats. Mol Cell Biol 32:2170–82.
  31. 31. Jimenez G, Ish-Horowicz D (1997) A chimeric enhancer-of-split transcriptional activator drives neural development and achaete-scute expression. Mol Cell Biol 17:4355–4362.
  32. 32. Gu T, Elgin SC (2013) Maternal depletion of Piwi, a component of the RNAi system, impacts heterochromatin formation in Drosophila. PLoS Genet 9:e1003780.
  33. 33. Li Y, Li H (2012) Many keys to push: diversifying the 'readership' of plant homeodomain fingers. Acta Biochim Biophys Sin (Shanghai) 44:28–39.
  34. 34. Wang W, Yu H, Long M (2004) Duplication-degeneration as a mechanism of gene fission and the origin of new genes in Drosophila species. Nat Genet 36:523–527.
  35. 35. Chauhan C, Zraly CB, Parilla M, Diaz MO, Dingwall AK (2012) Histone recognition and nuclear receptor co-activator functions of Drosophila cara mitad, a homolog of the N-terminal portion of mammalian MLL2 and MLL3. Development 139:1997–2008.
  36. 36. Bhaskar V, Courey AJ (2002) The MADF-BESS domain factor Dip3 potentiates synergistic activation by Dorsal and Twist. Gene 299:173–84.
  37. 37. St. Pierre SE, Ponting L, Stefancsik R, McQuilton P, FlyBase Consortium (2014) FlyBase 102 - advanced approaches to interrogating FlyBase. Nucleic Acids Res 42(D1):D780–D788.
  38. 38. Frydrychova RC, Mason JM, Archer TK (2008) HP1 is distributed within distinct chromatin domains at Drosophila telomeres. Genetics 180:121–31.
  39. 39. Park DJ, Pask AJ, Huynh K, Renfree MB, Harley VR, et al. (2004) Comparative analysis of ATRX, a chromatin remodeling protein. Gene 339:39–48.
  40. 40. Drané P, Ouararhni K, Depaux A, Shuaib M, Hamiche A (2010) The death-associated protein DAXX is a novel histone chaperone involved in the replication-independent deposition of H3.3. Genes Dev 24:1253–65.
  41. 41. Lewis PW, Elsaesser SJ, Noh KM, Stadler SC, Allis CD (2010) Daxx is an H3.3-specific histone chaperone and cooperates with ATRX in replication-independent chromatin assembly at telomeres. Proc Natl Acad Sci USA 107:14075–14080.
  42. 42. Clynes D, Higgs DR, Gibbons RJ (2013) The chromatin remodeller ATRX: a repeat offender in human disease. Trends Biochem Sci 38:461–6.
  43. 43. England BP, Admon A, Tjian R (1992) Cloning of Drosophila transcription factor Adf-1 reveals homology to Myb oncoproteins. Proc Natl Acad Sci USA 89:683–687.
  44. 44. Casola C, Lawing AM, Betrán E, Feschotte C (2007) PIF-like transposons are common in Drosophila and have been repeatedly domesticated to generate new host genes. Mol Biol Evol 24:1872–1888.
  45. 45. Maheshwari S, Wang J, Barbash DA (2008) Recurrent positive selection of the Drosophila hybrid incompatibility gene Hmr. Mol Biol Evol 25:2421–2430.
  46. 46. Jung SH, Evans CJ, Uemura C, Banerjee U (2005) The Drosophila lymph gland as a developmental model of hematopoiesis. Development 132:2521–2533.
  47. 47. Minakhina S, Steward R (2006) Melanotic mutants in Drosophila: pathways and phenotypes. Genetics 174:253–263.
  48. 48. Minakhina S, Tan W, Steward R (2011) JAK/STAT and the GATA factor Pannier control hemocyte maturation and differentiation in Drosophila. Dev Biol 352:308–16.