- Research article
- Open Access
Genomic structure and expression of Jmjd6 and evolutionary analysis in the context of related JmjC domain containing proteins
© Hahn et al; licensee BioMed Central Ltd. 2008
Received: 22 June 2007
Accepted: 18 June 2008
Published: 18 June 2008
The jumonji C (JmjC) domain containing gene 6 (Jmjd6, previously known as phosphatidylserine receptor) has misleadingly been annotated to encode a transmembrane receptor for the engulfment of apoptotic cells. Given the importance of JmjC domain containing proteins in controlling a wide range of diverse biological functions, we undertook a comparative genomic analysis to gain further insights in Jmjd6 gene organisation, evolution, and protein function.
We describe here a semiautomated computational pipeline to identify and annotate JmjC domain containing proteins. Using a sequence segment N-terminal of the Jmjd6 JmjC domain as query for a reciprocal BLAST search, we identified homologous sequences in 62 species across all major phyla. Retrieved Jmjd6 sequences were used to phylogenetically analyse corresponding loci and their genomic neighbourhood. This analysis let to the identification and characterisation of a bi-directional transcriptional unit compromising the Jmjd6 and 1110005A03Rik genes and to the recognition of a new, before overseen Jmjd6 exon in mammals. Using expression studies, two novel Jmjd6 splice variants were identified and validated in vivo. Analysis of the Jmjd6 neighbouring gene 1110005A03Rik revealed an incident deletion of this gene in two out of three earlier reported Jmjd6 knockout mice, which might affect previously described conflicting phenotypes. To determine potentially important residues for Jmjd6 function a structural model of the Jmjd6 protein was calculated based on sequence conservation. This approach identified a conserved double-stranded β-helix (DSBH) fold and a HxDxnH facial triad as structural motifs. Moreover, our systematic annotation in nine species identified 313 DSBH fold-containing proteins that split into 25 highly conserved subgroups.
We give further evidence that Jmjd6 most likely has a function as a nonheme-Fe(II)-2-oxoglutarate-dependent dioxygenase as previously suggested. Further, we provide novel insights into the evolution of Jmjd6 and other related members of the superfamily of JmjC domain containing proteins. Finally, we discuss possibilities of the involvement of Jmjd6 and 1110005A03Rik in an antagonistic biochemical pathway.
The jumonji C (JmjC)-domain containing proteins are an extending family of redox enzymes that catalyse a wide range of oxidation reactions. These proteins are found in all living organisms and are characterised by sharing the highly conserved JmjC domain. This domain was first described by Takeuchi and colleagues, who isolated a gene in the mouse which they called jumonji (meaning cruciform in Japanese), in regard to an abnormal phenotype visible in mutant embryos during neural plate development . Since then, the JmjC domain has found its entry into public domain databases and many other proteins containing this domain have been annotated or identified subsequently [2–5]. Currently, more then 11,000 sequence entries of JmjC domain containing proteins are present in Uniprot, PFAM, Interpro and SMART, thus demonstrating the tremendous expansion of members in this protein superfamily. Structural studies showed that the JmjC domain has a very characteristic topology. It forms a double-stranded β-helix (DSBH) fold, also known as jelly-roll motif, that normally consists of eight antiparallel β-strands . This DSBH topology is characteristic for the cupin metalloenzymes  and for the JmjC domain containing proteins . Cupin metalloenzymes constitute a large and diverse superfamily of proteins with enzymatically active and non-active members. They contain at least one DSBH or jelly-roll motive. Secondary structure elements that surround the DSBH are used further to define different subfamilies (for review see ). The DSBH fold in JmjC domain containing proteins can form an enzymatically active pocket by coordinating di-iron (Fe(II)) and the co-substrate 2-oxoglutarate (2OG). In almost all cases, the two-electron oxidation of the "prime" substrate is coupled to the conversion of 2OG into succinate and CO2 . Therefore, this class of enzymes is also known as nonheme-Fe(II)-2-oxoglutarate-dependent dioxygenases.
Oxidative reactions catalysed by 2OG-dependent dioxygenase are critical steps in the biosynthesis of metabolites, or post translational modification of interacting target proteins involving hydroxylation or demethylation and DNA/RNA repair of N-methylated nucleic acids [6, 9]. As such, 2OG-dependent dioxygenases are of immanent biological importance. A number of JmjC domain containing proteins are involved in regulating key developmental processes during mammalian embryogenesis by acting as transcription factors. An example of this subgroup of JmjC domain containing proteins is jumonji itself (now called Jarid2), which can act as a transcriptional repressor. Jarid2 inhibits cyclin D1 transcription during heart and brain development and thereby controls cellular proliferation and differentiation in these organs during morphogenesis [10–12]. Other examples of JmjC-domain proteins that function as transcriptional repressors are the retinoblastoma-binding protein 2 (Jarid1a, also known as RBP2), which is required for the differentiation of various cell types including muscle, bone and myeloid cells  and hairless, a nuclear co-repressor of multiple nuclear hormone receptors such as the vitamin D receptor, the thyroid hormone receptor and the retinoic acid-receptor-related orphan receptor α [14–16]. Mutations in hairless result in humans and mice in alopecia, a form of hereditary hair loss due to defects in the regeneration of hair follicle cells [17–19]. More recently, JmjC-domain containing proteins have gained emerging importance in histone modification and chromatin regulation. The JmjC domain containing histone demethylases (JHDMs) can catalyse lysine demethylation of histones through an oxidative reaction that requires iron Fe(II) and 2OG. Histone methylation marks have important roles in regulating gene expression and are central to the control of epigenetic information that regulates cell fate and identity . Until now, JHDMs have been reported to be capable to reverse H3K36 (JHDM1/FBXL11, , JMJD2A, [22, 23], Rph1, [24, 25], Jhd2, , Gis1, , Jhd1, ), H3K9 (JHDM2A/JMJD1A, ), and both H3K9 and H3K36 methylation (JHDM3A to JHDM3D/JMJD2A to JMJD2D, [23, 27–30]). In addition, the JARID subfamily of JmjC-domain containing proteins have been shown to possess H3K4 demethylase activity in different species (JARID1A/RBP2, [31, 32], JARID1C/SMCX, [31, 33], JARID1D/SMCY, , JARID1B/PLU-1/Yjr119Cp/Jhdp2, [31, 35, 36], Lid, [26, 37–39], and JMJ2 ). Most recently, the JmjC-domain containing proteins UTX and JMJD3 have been shown to function as H3K27 specific demethylases and demonstrated to play an important role in animal body patterning and regulation of inflammatory responses [41–44]. Through this activity the JHDMs have been postulated to regulate diverse cellular processes including control of cell proliferation, heterochromatin assembly and gene expression [26, 28, 30, 34, 45, 46]. However, for most of the newly discovered JHDMs gene ablation studies in vivo have as yet not been carried out leaving the possibility open that in addition to histones other substrates might be targets of these proteins.
Many of the different classes of JmjC domain containing proteins possess complex domain architectures by having at least one additional protein domain besides the JmjC domain. Here, the JHDMs and Jarid proteins are good examples. In addition to the JmjC domain they have PHD, Tudor, Bright/Arid and zinc finger domains, which have recently been shown to interact with methylated histone tails and are required for their catalytic function [33, 38, 47]. In contrast, other members in this protein superfamily have no other recognisable protein domain apart from the JmjC domain. Important members of this subgroup of JmjC domain only proteins are the 2OG-dependent dioxygenase factor inhibiting HIF (Hif1an, also known as FIH) and the PHD hydroxylases, which are involved in the maintenance of oxygen homeostasis. Hif1an was the first structurally resolved mammalian JmjC domain containing protein [48, 49] and was shown to function as an asparagine hydroxylase for the hypoxia inducible factor α (HIF1α), a transcription factor that is central to oxygen homeostasis in both physiological and pathophysiological processes (for review see ). The PHD hydroxylases PHD1 to PHD3 also target HIF1α by hydroxylating proline residues. Under normal oxygen conditions the hydroxylation reactions carried out by PHDs induce the proteasomal degradation of the HIF1α subunit, whereas under hypoxia HIF1α is not hydroxylated and translocates from the cytoplasm to the nucleus where it activates transcription of genes needed for adaptation to low oxygen levels. AlkB is another member of the JmjC domain only proteins. As a 2OG-dependent oxygenase AlkB can repair DNA that is methylated at 1-methyladenine and 3-methylcytosine by oxidation of the methyl group [51, 52]. AlkB and related proteins define a subfamily of small 2OG-dependent oxygenases that are involved in DNA and RNA repair .
In the past, we have analysed the function of the phosphatidylserine receptor (Ptdsr or also known as PSR or PtdSerR), a gene that also encodes a JmjC domain only protein [54, 55]. First described as a cell surface receptor for the recognition and engulfment of apoptotic cells , recent studies have raised serious doubts about its involvement in apoptotic cell clearance. Five groups independently reported a nuclear localisation of Ptdsr, which clearly contradicts the hypothesis that the protein might function as a transmembrane receptor [57–61]. Furthermore, a comprehensive analysis carried out by us in Ptdsr-knockout mice demonstrated no impairment in the removal of apoptotic cells in Ptdsr-deficient embryos and macrophages ( for Ptdsrtm 1Gbf) thereby contradicting two previous studies that reported defects in apoptotic cell clearance in two other Ptdsr-knockout mice ( for Ptdsrtm 1Flv, and  for Ptdsrtm 1Ysfk). All studies involving Ptdsr-knockout mice conclusively demonstrated that the gene is essential for normal development. Ptdsr-deficient mice are dying at birth and display severe differentiation defects in many tissues such as brain, eyes, lung, kidney, liver and intestine at different stages of embryogenesis [54, 62, 63]. Furthermore, Ptdsr plays an important role during heart development and ablation of Ptdsr function is associated with complex cardio-pulmonary malformations that resemble the human congenital heart syndrome Tetralogy of Fallot .
Although Ptdsr is clearly an important differentiation-promoting gene during embryogenesis its molecular function is still unknown. Based on a bioinformatic analysis of the cloned Hydra homologue PSR, Cikala and colleagues were the first to suggest that the protein might function as a 2OG-dependent dioxygenase that is capable of modifying unknown nuclear targets . In recognition of the still unknown function of the "phosphatidylserine receptor gene" the International Committee for Standardized Genetic Nomenclature in Mice (ICSGNM) has most recently revised its name. Ptdsr has been renamed into jumonji domain containing gene 6 (Jmjd6). We will obey the guidelines of the ICSGNM and use this name and the gene symbol Jmjd6 throughout the remainder of this article.
To gain more insights into possible Jmjd6 functions using sequence homology approaches, we developed a semiautomated process to extract all available Jmjd6 homologous protein and nucleic acid sequences from public databases. This comprehensive data set was used to identify proteins with sequence homology to Jmjd6 in 62 species across all major living phyla. In addition, this allowed us to identify new Jmjd6/JMJD6 splice variants in mice and humans and to analyse the genomic neighbourhood of the gene in vertebrates. Immediately downstream of Jmjd6 we found the gene 1110005A03Rik, which is predicted to encode a putative methyltransferase, which is as yet not further biochemically characterised. Interestingly, we found 1110005A03Rik to be affected by deletions in two previously reported Jmjd6 knockout mice. To predict possible structure-function relationships of Jmjd6, we employed a comparative modelling approach of the three-dimensional structure of the protein and annotated probable functional residues based on sequence conservation. To integrate Jmjd6 into the superfamily of JmjC-domain containing proteins, we assembled a non-redundant set of 313 DSBH-domain containing proteins by using our semiautomatic retrieval script and grouped these according to phylogenetic relationships. Taken together, our results further support a conserved enzymatic function of Jmjd6 most likely as a classical non-heme, Fe(II)- and 2OG-dependent dioxygenase. It also shows that Jmjd6 is closely related to Jmjd4 and JFP6, and constitutes together with these two proteins a subgroup of JmjC domain containing proteins. Moreover, we identified in protista, fungi and plants proteins with sequence homology to Jmjd6 that diverge from the classical "JmjC-only domain" structure. Some of these proteins have in comparison to the animalia Jmjd6 extended C-terminal regions that contain additional domains that might have been lost during evolution.
With these approaches we establish here an overall framework of information that will facilitate not only further research regarding the functional characterisation of Jmjd6 but also of other members within the DSBH-domain containing superfamily of proteins.
Retrieval and analysis of a comprehensive set of Jmjd6 orthologs and other JmjC domain containing proteins
We started the "DNA" sequence based retrieval and analysis approach (depicted in yellow in Figure 1) by identification of all mouse Jmjd6 homologous sequences available in Ensembl (release 44). This candidate loci list contained homologous Jmjd6 loci in fully annotated genomes as well as Jmjd6 sequence fragments from ongoing sequencing projects. To identify conserved coding and non-coding sequences within these genomic regions we undertook a PipMaker analysis of retrieved sequences that showed a similar genomic organisation in comparison to the murine Jmjd6 locus. Next, we mapped all available murine and human ESTs in dbEST to the murine and human Jmjd6/JMJD6 loci, respectively, using the program e2g. Extensive validations of each EST employing Sim4 with our own perlscript for data retrieval, handling, examination and sorting of the results led to a comprehensive analysis of the murine and human Jmjd6/JMJD6 loci and a detailed annotation of all encoded transcripts.
To identify all Jmjd6 orthologs, we employed a blastp search against Ensembl and nr databases using a unique N-terminal sequence of the murine Jmjd6 protein as a template ("protein" based search and analysis, depicted in Figure 1 in green). The candidate matches were individually verified in a Mus musculus restricted blastp database search using the reciprocal best BLAST hit criterion and a perlscript for automatic data processing. These led to the identification of Jmjd6 proteins in other species that are closely related to and might be orthologs of the murine Jmjd6 protein. A full length ClustalW alignment of proteins with significant homology to Jmjd6 was processed using a TeXshade based conservation analysis and also as input for a 3D structure prediction of the catalytic core domain. By combining these results, we developed a novel way of predicting residues that are most likely essential for the catalytic activity of the protein based on their degree of conservation.
To analyse the phylogenetic evolution of the Jmjd6 JmjC domain, we developed SMART-parser. SMART and SMART-parser were used to analyse the domain composition and extract the sequence information from a set of sequences that was derived from the Ensembl, Interpro, Uniprot and SMART databases ("DSBH-domain" routine, depicted in Figure 1 in blue). This set comprises all JmjC-, Pfam-JmjC, Blast-JmjC-, P4Hc-, Pfam-2OG-FeII-Oxy- and Pfam-PhyH-domain containing sequences from human, mouse, zebrafish, pufferfish, nematode, fruit fly, and yeast. These different domain names have been used in the past to annotate the JmjC domain in DSBH fold containing proteins. The SMART-parser verified sequences were used as an input for a global ClustalW alignment. A PHYLIP processing of all ClustalW alignments allowed the analysis of the Jmjd6 phylogeny in relationship to all other identified DSBH domain containing proteins.
Comparative analysis of Jmjd6 homologous loci in vertebrates
Species distribution and chromosomal locations of Jmjd6 orthologs in mammalian and non-mammalian genomes
Cluster with 1110005A03Rik
Cluster with Sfrs2
Gene scaffold 591
Western clawed frog
Super contig 1.887
Based on the identification of the highly conserved sequence coding for the new exon 5 in all mammalian species (Figure 2 and data not shown), we performed a similar e2g and Sim4 analysis using the human JMJD6 locus and human EST database entries. This analysis confirmed the presence of exon 5 containing JMJD6 transcripts in humans and also the presence of two different variants of exon 4 in the human genome. Using again RT-PCR, we validated the expression of these transcripts in vivo. Similar to the mouse we found expression of the three transcripts in 10 human tissues [see Additional file 1]. These, as well as the mouse alternative transcripts, were further validated and confirmed by sequencing (Accession numbers EF527404 for Mm2 and EF527405 for Mm3 for mouse Jmjd6 and EF527406 for Hs2 and EF527407 for Hs3 for human JMJD6 transcript variants, respectively). In comparison to the new identified mouse transcript variants, both human transcript variants are predicted to generate truncated proteins. Neither of these predicted proteins appear to encode a nuclear export signal, sumoylation site, AT hook domain or poly-serine stretch at the C-terminal part of the protein (Fig. 3C). As an insertion of premature translation termination codons might result in nonsense-mediated decay (NMD) of the corresponding mRNA , we analysed if the mouse transcript variant 2 and human splicing variants 2 and 3 might be subjected to degradation by the NMD pathway. Using RT-PCR assays for each of these variants we found that this is not the case. Primers located in the new introduced exons and at the 3'-end of each transcript yielded products of the expected size, indicating that these variants are indeed expressed and most likely functional [see Additional file 2]. The human and mouse splice variants seem to be expressed at different mRNA levels depending on the tissue analyzed (see Figure 3 and [see Additional file 2]), with mouse splice variant Mm1 expressed in higher amounts in the developing embryo (Figure 3). Interestingly, an e2g and Sim4 analysis performed in non-mammalian species did not result in the recognition of mouse and human related Jmjd6 splice variants in these organisms (data not shown).
We then went on to investigate the consequences on Jmjd6, 1110005A03Rik and Sfrs2 expression of the different gene targeting strategies used to generate the three Jmjd6 mouse knockout alleles. In all three Jmjd6 mutant mice (Jmjd6tm 1Flv, generated by the Flavell laboratory; Jmjd6tm 1Ysfk, generated by the Fukui laboratory; and Jmjd6tm 1Gbf, generated in our laboratory) the first two or three exons of the gene have been deleted. However, an alignment of restriction enzyme sites used to construct the different gene targeting vectors reveals that in Jmjd6tm 1Flvand Jmjd6tm 1Ysfkmice, in contrast to the Jmjd6tm 1Gbfmouse line, exon 1 of 1110005A03Rik was deleted in addition [see Figure A in Additional file 3]. This was confirmed when we analysed the expression of all three genes in homozygous embryos harvested from the three mutant mouse lines. Jmjd6tm 1Flv-/-and Jmjd6tm 1Ysfk-/-animals show no expression of 1110005A03Rik and Jmjd6, while the expression of Sfrs2 remains unaffected by the gene targeting. In Jmjd6tm 1Gbfhomozygous embryos 1110005A03Rik, and Sfrs2 are expressed comparable to wild type levels [see Figure B in Additional file 3] and as previously shown , only the expression of Jmjd6 is ablated in this mouse mutant line. This was further supported by semiquantitative expression analysis of 1110005A03Rik in Jmjd6tm 1Gbf-/-embryos and primary fetal liver-derived macrophage cultures. RT-PCR analysis of serial diluted cDNA shows equal expression of 1110005A03Rik in Jmjd6tm 1Gbfanimals in comparison to wildtype controls in all analysed samples [see Figure C in Additional file 3]. Thus, the Jmjd6tm 1Flvand Jmjd6tm 1Ysfkalleles represent double knockouts of the Jmjd6 and 1110005A03Rik genes, while the Jmjd6tm 1Gbfallele ablates only Jmjd6 gene expression.
Phylogenetic analysis of the Jmjd6 protein
To explore potentially important residues for Jmjd6 function we compared the degree of sequence conservation across the aligned sequences and included in the alignment predicted tertiary structure elements of the protein. First, our analysis showed the presence of a 2-His-1-carboxylate facial triad (HxDxnH), as a common structural motif among non-heme iron(II) enzymes, present in all identified Jmjd6 proteins (Figure 5), suggesting an evolutionary conserved enzymatic function of Jmjd6. We could identify within the predicted JmjC domain eight β-strands (I-VIII) that have been shown to form the characteristic jelly-roll topology in other structurally resolved JmjC domain containing proteins . The amino acids encompassing these β-strands are highly conserved in all proteins. 50% of all β-strands forming amino acids are identical in at least 85% of the analysed species suggesting the spatial and functional conservation of the JmjC-domain in all analysed proteins. The only exception we found is in between β-strands IV and V. At this position we found an insertion of sequences of variable amino acid conservation among species and of varying length as it has been previously described for other JmjC domain containing proteins [6, 22, 53]. Second, within the C-terminal region of the Jmjd6 protein preceding the last β-strands of the recognised jelly-roll the degree of sequence conservation is decreased. In contrast to the metazoan species showing a high degree of sequence identity, all other species display a significantly reduced similarity in this region. Furthermore, only the metazoan species have a poly-serine stretch in the C-terminal region of the Jmjd6 protein (Figure 5 and ). We could not identify any comparable sequence motif in the other species but looking for chemical properties of amino acids present at this position in proteins of non-metazoan species we found an enrichment of polar amino acids in this part of the protein (data not shown). This suggests at least a conservation and fortification of physico-chemical properties in this region during evolution. Third, the part N-terminal of the JmjC domain shows in comparison to the C-terminus a higher degree of conservation. We found that 20% of the amino acids 59–140 are identical in at least 90% of the analysed proteins (Figure 5). In addition, we identified one highly conserved and three less conserved β-strands in this region. These findings suggest the presence of an important structural element of the Jmjd6 protein in close vicinity to the predicted JmjC domain (see below).
Comparative modelling and analysis of the conserved putative catalytic core domain of Jmjd6
Following iterative backbone annealing based on the structure of Hif1an (PDB entry 1h2k), 2OG and Fe(II) were added to the 3D model of Jmjd6. After energy minimisation utilising the SwissPDB Viewer, the predicted active site of the Jmjd6-Fe(II)-2OG complex was analysed in detail. Here, we found the proper positioning of the 2-His-1-carboxylate facial triad chelating the iron atom in a manner similar to other 2OG- and Fe(II)-dependent dioxygenases (Figure 6B). The HxD motif (His 187 and Asp 189) is predicted to be located just after the sequence of DSBH β-strand II and the second Fe(II)-binding histidine (His 273) is located on DSBH β-strand VII. The observation of Fe(II) coordination by these three residues confirms their initially suggested assignment based on the primary sequence analysis of the multiple sequence alignment of Jmjd6 proteins (Figure 5). In addition, we could identify in the model several residues within the putative active core of Jmjd6 capable of positioning the co-substrate 2OG in the cavity formed by the interior of the DSBH. The proper positioning of 2OG relative to the Fe(II) seems to be achieved by an interaction of several residues of the active core that are close enough to the hydrogen-bond with the 2OG. Among these residues we predict Asn197 and Trp174 to interact with the 1-carboxylate oxygen of 2OG that is not ligated to the Fe(II). Asn197 and Trp174 are stabilised by Thr285. Interestingly, in Jmjd6, like in Hif1an and Jmjd2a, a lysine (Lys204, from DSBH β-strand IV) could be important for the binding of the C5'-carboxylate of 2OG. Most of the other structurally resolved 2OG- and Fe(II)-dependent dioxygenases have instead of the threonine 285 an arginine at this position, which is part of a RXS motif within the DSBH β-strand VIII [76, 77]. Thus, Jmjd6 has in contrast to other 2OG-dependent dioxygenases no tripeptide RXS motif (Figure 5). In addition to these major 2OG interacting residues in Jmjd6, we could recognise additional residues (Phe133 and Val275) for putative hydrophobic interactions in the active core of the protein.
Integration of Jmjd6 into the JmjC domain containing protein family
Interestingly, in some subgroups we observed a huge expansion in the number of family members. The P4H/Phy-3 family of prolyl-4-hydroxylases contained 22 proteins from Drosophila melanogaster. The P4H/Phy-3 subgroup compromises prolyl-4-hydroxylases involved in collagen biosynthesis or hypoxia sensing. Currently, it is unknown why especially the fly has so many members in this protein family. Another example is the AlkB subfamily of DNA and RNA demethylating enzymes. This group contains 45 proteins that split into the Alkbh1 to Alkbh8 protein family members. Apart from Alkbh1 to Alkbh3 that are known to be DNA/RNA repair enzymes [52, 78, 79] the exact biochemical activity of most of the other subfamily members has still to be determined.
Following phylogenetic analysis and tree generation a subtree encompassing the Jmjd6 protein was extracted for further analysis (Figure 8). By combining the information from the phylogenomic analysis with the annotation of protein subfamilies it is evident that Jmjd4 is closely related to Jmjd6 regarding sequence similarity. The JmjC domain of both proteins shows the highest sequence similarity among all 313 classified members (32% sequence identity and 50% sequence similarity). This indicates that Jmjd4 is the closest paralogue of Jmjd6.
Moreover, a detailed sequence analysis of JmjC domain containing protein subfamilies that are related to the Jmjd6 and Jmjd4 subgroups revealed further insights into Fe (II) and 2OG coordinating mechanisms (Figure 8). The JFP6 subgroup has a HxHxH motif instead of the Hx(D/E)xnH motif that is commonly used by all other JmjC domain containing proteins for the coordination of Fe(II) (Figure 8, orange box). If this motif is compatible with an oxygenase activity, as one would expect, has to be investigated in the future. Furthermore, the Jmjd5 homologue of Tetraodon nigroviridis, all Phf2 proteins, and Epe1 have only a Hx(D/E) motif (red box in Figure 8) and therefore lack the second histidine, which is known to be important for iron binding. Moreover, the T. nigroviridis Phf2 homologue lacks in addition the lysine that stabilises the C5'-caboxylate of 2OG. Next to this motif, Tuskada et al. described residues in positions -3, +6–8 and +13–15, in respect to the HxD motif to be important for the catalytic function of histone demethylases . In the -3 position, the 11100034B05Rik, JFP6 and No66/Mina subgroups have a nonpolar side chain present in contrast to the polar side chains of serine/threonine residues in other subgroups. In the +6–8 position, members of the Pla2g4b, FBXL10/FBXL11, Phf2/Phf8, and individual members of the JFP6 and Jhd1 subgroup have a conserved tyrosine. In Epe1, this tyrosine has been shown to be important for heterochromatin destabilisation, which is independent of an Epe1 histone demethylation activity [21, 46]. Another interesting feature is the predicted presence of an extended major β-sheet in most of the JmjC domain only protein families encompassing the subtree (Figure 8).
Here we report a systematic annotation and comparison of Jmjd6 homologous proteins across diverse phyla. With the recent description of the JmjC domain as a novel demethylase signature motif  and the discovery that many proteins containing JmjC domains are involved in reversing histone methylation marks [4, 80], this protein superfamily has attracted central interest in many different fields of biology. It has become quite clear that these enzymes have not only important functions in the epigenetic regulation of chromatin modification states but also in diverse other physiological contexts ranging from oxygen sensing to fatty acid metabolism and DNA repair . However, for most of these proteins it has remained difficult to unravel their molecular function or to predict their enzymatic targets solely based on sequence comparisons and analysis of their JmjC domains. In the past, we have been interested in the "phosphatidylserine receptor" (now renamed into jumonji domain containing 6), a JmjC domain containing protein, which was initially described to function as a type II transmembrane receptor for the engulfment of apoptotic cells . To compare the JmjC domain of Jmjd6 to the DSBH/JmjC domains of other members within this protein superfamily we designed a semiautomated sequence retrieval and analysis routine that allowed us to establish a catalogue of 313 DSBH/JmjC domain containing proteins in nine selected species. In addition, we analysed the Jmjd6 and neighbouring loci in vertebrates and investigated their modes of expression.
Comparative analysis of the vertebrate Jmjd6 locus and immediate neighbouring loci
With the ongoing genome sequencing projects in many different species and integrated efforts to annotate the functional content of fully sequenced genomes there is a need to complement existing automated genome annotations with a focused bioinformatics analysis of individual genes and gene families to make fully use of publicly available databases and resources. Such approaches might help to correct mistakes in automated annotation procedures and might add additional information to gene sets. This will finally lead to a more comprehensive and accurate annotation of gene functions.
We used the PipMaker alignment tool to analyse the patterns of sequence conservation across the mouse Jmjd6 and the neighbouring loci 1110005A03Rik and Sfrs2. To identify all protein coding exons as well as highly conserved non-coding sequence elements in this region we analysed genomic sequences of 22 different vertebrate species. This led to the identification of an up to now overseen additional Jmjd6 exon in mammals and to the detection of a conserved putative bi-directional promoter element in between the closely spaced Jmjd6 and 1110005A03Rik genes. The re-annotated genomic region encompassing the Jmjd6, 1110005A03Rik and Sfrs2 loci was then the basis for a comparison of gene-targeting strategies that have been used to generate three different loss-of-function alleles of Jmjd6 in the mouse [54, 62, 63]. By aligning a map of restriction enzyme sites used for the construction of the different targeting vectors with the genomic region encompassing the three genes, we could show that in the Jmjd6tm 1Flvallele and in the Jmjd6tm 1Ysfkallele exon 1 of the 111005A03Rik gene as well as the conserved intergenic region in between Jmjd6 and 111005A03Rik was deleted upon homologous recombination in ES cells. An expression analysis of Jmjd6, 111005A03Rik and Sfsr2 in the three different mouse alleles confirmed that Jmjd6tm 1Flvand Jmjd6tm 1Ysfkmutant mice are double knockouts of the Jmjd6 and 111005A03Rik genes. In contrast, the Jmjd6tm 1Gbfallele ablates only the expression of Jmjd6. Although all three independently generated knockout mice show to some degree overlapping phenotypes regarding defects during embryogenesis (e.g. disturbed eye development, brain malformations and defects in haematopoiesis) other controversial observations concerning the description of specific phenotypes have been reported [54, 62, 63]. Foremost, the capabilities of the mutant mouse lines to engulf apoptotic cells have been described differently. While Kunisaki et al.  and Li et al.  reported that their knockout mice showed impairment in the clearance of apoptotic cells during embryogenesis and defects in apoptotic cell phagocytosis assays, no such differences in phagocytosis competence could be detected in Jmjd6tm 1Gbfmutant mice . It has been suggested that differences in genetic backgrounds might account for these reported phenotypic variations [82–84]. However, in the light of the described results with Jmjd6tm 1Flvand Jmjd6tm 1Ysfkmutant mice being double knockouts of Jmjd6 and 111005A03Rik the situation might be far more complex. It could well be that loss-of 111005A03Rik function in Jmjd6tm 1Flvand Jmjd6tm 1Ysfkknockout mice might contribute in addition to the Jmjd6-deficiency to the phenotypic discrepancies observed. Future studies involving the generation and analysis of 111005A03Rik knockout mice will address this possibility.
Interestingly, the as yet functionally uncharacterised 111005A03Rik gene contains a predicted arginine protein methyltransferase domain. We demonstrate here, that the conserved sequence element in the short intergenic region between Jmjd6 and 111005A03Rik possess bi-directional promoter activities for both genes. Of note, the forward orientation of the fragment, which controls 111005A03Rik expression in vivo showed only half of the transcriptional activity as compared to the reverse orientation, which controls Jmjd6 expression in vivo. This indicates that both genes might compose a bi-directional transcriptional unit with opposing transcriptional and regulatory activities. This gives room for speculations. If 111005A03Rik has a function as a methyltransferase and the Jmjd6 protein possess demethylation activity as many other JmjC domain containing proteins are doing, both proteins might be involved in a common biochemical pathway targeting possibly a shared substrate. Interestingly, such counterbalancing principles of methylation and demethylation activity have been already suggested for other JmjC domain containing proteins . For example, the Alkbh8 protein contains an amino-terminal JmjC domain, which might possess DNA demethylase activity that is fused to a predicted C-terminal methylase domain. This domain architecture is conserved in the homologous Alkbh8 proteins in Drosophila (CG17807) and C. elegans (C14B1.10) suggesting the selection of a controlled methylation/demethylation regulatory mechanism during the evolution of these proteins.
If Jmjd6 and 111005A03Rik also establish such counteracting functions needs to be addressed in the future. To proof such a hypothesis, the molecular function of both genes need first to be characterised. Nevertheless, the additional deletion of the 111005A03Rik gene in two of the three described Jmjd6 knockout mice might well explain differences reported for their apoptotic cell engulfment phenotypes. If 111005A03Rik is functioning as a methyltransferase in the control of transcriptional regulation, it is possible that the expression of other genes targeted or regulated by 111005A03Rik and/or Jmjd6 might influence phagocytosis of apoptotic cells. In such a scenario the described engulfment defects might be related to secondary effects caused by deregulation of genes involved in processes of phagocytosis or cellular differentiation of phagocytes.
Using PipMaker analysis based on sequence conservation we identified a before overseen exon in Jmjd6 loci of mammalian species. A subsequently undertaken e2g annotation of all available EST sequences to the Jmjd6/JMJD6 genomic regions lead to the discovery of two new alternative transcripts in mice and humans. At the protein level the most distal C-terminal part of the Jmjd6 protein (last 100 amino acids) is affected in all new identified transcripts. Either truncated polypeptides are generated that miss the poly-serine stretch, the putative AT-hook domain, and the predicted sumoylation and nuclear export signal (NES) or a stretch of eight amino acid is exchanged with 31 other residues. In this second splice variant identified in mice (Mm3) the predicted sumoylation site, the NES and the poly-serine stretch are retained but the putative AT-hook domain is missing (Figure 3C). Our 3D structural model of the Jmjd6 protein indicates that the sumoylation site, the NES and the AT-hook motif might be accessible for interacting proteins. The AT-hook domain is known to mediate binding to DNA in high-mobility group proteins . Jmjd6 is a nuclear protein, which has been shown to contain five independently active nuclear localisation signals . The NES might allow a shuttling of the protein from the nucleus to the cytoplasm. Likewise, the sumoylation site might be used for the regulation of interactions with other proteins.
A comparison of the Jmjd6 sequence with the resolved protein structure of the Hif1an/HIF1α C-terminal transactivation domain co-complex indicates that interactions with other proteins might occur in the Jmjd6 C-terminal region . Alternatively, at this site homo-di- or homo-multimerization of Jmjd6 polypeptides might take place. For many JmjC/DSBH domain containing enzymes it is known that they can exist in various oligomeric forms. For example, Hif1an is a dimer in solution, germin forms a hexamer and HPPE a tetramer (for review see ). Indeed, very recently it could be demonstrated that Jmjd6 forms covalent multimers through homomeric polymerisation of the 50 kD protein . Nuclear aggregated Jmjd6 products with masses of 100, 150, 200 and 250 kD were detected after immunoprecipitation indicating that assembly of homomultimers might be important for the protein function . Thus, the new alternative transcripts encoded by the Jmjd6 locus might possess a regulatory role by suppressing multimer formation or interactions with other proteins. Such possibilities need to be experimentally investigated in the future.
Structural model of the Jmjd6 protein and evolutionary relationship to other JmjC/DSBH-domain containing proteins
Recent crystallographic studies have helped to provide important insights in the structure and functions of JmjC domain containing proteins. More than ten structures of different members of this large oxygenase family have been solved demonstrating that the JmjC domain forms a characteristic DSBH structural fold composed of eight β-strands that build two β-sheets each containing four anti-parallel β-strands . The catalytic core of these metalloenzymes consists of a highly conserved Hx(E/D)xnH signature motif that supplies three chelating positions for the Fe(II) iron and two additional chelating positions are contributed by the substrate co-factor 2OG. Very often the binding sites for 2OG in the DSBH core are as well conserved among the members of this protein family. Due to high sequence similarity of Jmjd6 to other JmjC domain containing proteins , we have embarked on the modelling of the Jmjd6 3D-structure to gain further insights into possible protein functions. Our analysis shows that Jmjd6 is indeed most likely a DSBH-fold containing 2OG-dependent dioxygenase enzyme, as previously first suggested by Cikala and colleagues . All residues needed for Fe(II) coordination, 2OG co-substrate binding and oxidative catalysis were found to be located within the predicted DSBH domain of the protein model and are evolutionary preserved in all analysed Jmjd6 homologous proteins probably including those of non-metazoan species. We identified in all orthologous Jmjd6 proteins from metazoan species the characteristic HxDx[n]H catalytical triad and our model predicts five conserved residues (Trp174, Thr184, Asn197, Lys204, and Thr285) as being important for interaction with a 2OG co-substrate. In addition, the extreme conservation of residues forming the eight β-strands of the DSBH domain suggests a spatial and functional conservation of the Jmjd6 JmjC domain during evolution. This makes it highly likely that all Jmjd6 proteins in the different animal kingdoms are indeed active 2OG-dependent dioxygenase enzymes. This might not be true for other members of the JmjC/DSBH fold superfamily of proteins. For example, the subfamily of JFP6-like proteins has a substitution in the Fe(II) coordinating residues changing the Fe(II) binding motif. Furthermore, hairless and Jarid2 proteins are most likely enzymatically inactive due to the missing HxD motifs in their DSBH protein cores . This suggests that JmjC domains might possess additional functions that are independent of any enzymatic activity.
2OG-dependent dioxygenases are known to catalyse two types of posttranslational protein modifications. Described have been hydroxylation of the amino acid side chains of proline, asparagines, aspartic acid and tryptophan, and secondly demethylation reactions on lysine residues (for review see [6, 81]). It has been proposed that also methylated arginines might be targets of JmjC domain containing proteins. In fact, most recently it has been suggested that Jmjd6 might possess an enzymatic activity as an arginine histone demethylase. It was reported that Jmjd6 is capable to catalyse the removal of di-methyl groups on histone H3 at arginine 2 (H3R2) and on histone 4 at arginine 3 (H4R3) . Asymmetric di-methylation of H3R2 has been associated with repression of transcriptional activity through its enrichment in heterochromatic loci and inactive euchromatic genes [87, 88]. Thus, Jmjd6 might have a function in gene activation by erasing repressive chromatin marks. But it is also not unlikely that additional, non-histone substrates are targeted by Jmjd6.
All types of posttranslational modifications catalysed by 2OG-dependent dioxygenases involve a common type of hydroxylation reaction. During the catalytic reaction, the splitting of molecular oxygen is coupled to the hydroxylation of the target substrate and the oxidative decarboxylation of 2OG to give succinate and CO2. In the case of demethylation, hydroxylation of a methyl group is followed immediately by deformylation producing formaldehyde as an additional reaction product. As noted before by others, it is impossible to predict the prime substrate of 2OG-dependent dioxygenase based on a primary sequence analysis of the JmjC domain [80, 81]. Even JmjC domain containing enzymes of the same subfamily such as the JHDM1A and the JHDM2A histone demethylases that both recognise H3K36 methylation have very limited primary or secondary structure similarity . This suggests a requirement for an extended substrate recognition mechanism that probably involves protein-protein interactions in larger protein complexes or for a necessity of very defined interaction sites in the catalytic core that are important for determining substrate selectivity . This is further supported by a number of recent studies that implicated other methyl-dependent protein domains like Tudor, MBT, PHD fingers and chromo domains as reader domains in the recognition of the histone code [90–94]. Jmjd6 contains a single JmjC domain and beside the predicted recognition sequences for nuclear in- and export, sumoylation and the AT-hook DNA-binding motif no other recognisable domain is present in the protein. Thus, the substrate specificity of Jmjd6 must be regulated by interaction with other proteins.
What is quite clear is that the putative enzymatic function of Jmjd6 as a 2OG-dependent dioxygenase is linked to its important role in development and morphogenesis, at least in vertebrates. All mouse mutants conclusively point to an essential role of Jmjd6 during embryogenesis were the gene is involved in the differentiation of multiple tissues and organs. Also in zebrafish, the morpholino knockdown of the jmjd6 orthologue resulted in severe defects in somite, brain, heart and notochord development . This may or may not involve secondary effects on apoptosis induction and cell corpse clearance. Unlikely is a direct involvement of jmjd6 in apoptotic cell recognition or engulfment. Given that many 2OG-dependent dioxygenases have been implicated in the regulation of multiple pathways, global effects on gene expression and metabolism are not surprising when their enzymatic functions get perturbed or ablated.
The function of Jmjd6 during development seems to be different in vertebrates and invertebrates. Inactivation of the orthologous genes in Drosophila and C. elegans were surprisingly not associated with embryonic lethality and no major developmental defects have been reported for the loss-of-function mutants [59, 96]. Interestingly, mutant flies that lack dPSR show inappropriate apoptosis during eye development . Major phenotypes were observed in gain-of-function experiments in which the dPSR gene was overexpressed during development. Male transgenic flies that expressed dPSR under ubiquitous promoters developed a genital rotation phenotype. dPSR overexpression in the wing resulted in various wing blade fusion phenotypes and driving gene expression in the eye induced a small rough eye phenotype. Further genetic analysis pointed to an involvement of dPSR in the head involution defective (hid) and c-Jun-NH2 terminal kinase pathways. These suggested that loss of dPSR enhances apoptosis in the developing eye, while overexpression of dPSR protects from apoptosis. Of note, the analysis of mutant flies did not reveal a role for dPSR in the phagocytosis of apoptotic cells during development. Altogether, these observations support a function of dPSR in cell survival and differentiation that might have parallels in the vertebrate system .
When we compared the protein sequences of vertebrate and invertebrate orthologous Jmjd6 proteins we found divergence in the N- and C-terminal regions of the protein. Interestingly, the C-terminal poly-serine stretch is either shorter or missing in the invertebrate species. Moreover, we could not identify an additional exon in the invertebrate orthologous Jmjd6 genes that would generate alternative transcripts with effects on the C-terminal part of the protein. Therefore, it could well be that differences in this region of the protein are linked in vertebrates with different physiological functions of Jmjd6 protein, possibly through interactions with other proteins. This is also supported by identification of proteins in plants, fungi and protista that show partial sequence homology to the animalia Jmjd6 protein (Figure 5 and [see Additional file 4]). These were identified through sequence search routines focussing on a conserved N-terminal sequence stretch outside of the highly conserved JmjC domain. Consequently, these proteins show in their N-terminal parts significant sequence homology to the mouse Jmjd6 protein, but are diverse to Jmjd6 in their C-terminal regions. Moreover, many of these proteins are longer then the animalia Jmjd6 proteins (up to 2.000 aa instead of around 400 aa for animalia Jmjd6 proteins) and contain often in addition to the JmjC domain a single F-box domain.
Thus, it is very likely that they possess different or additional functions than the animalia Jmjd6 proteins and a possible homology of these proteins to Jmjd6 should be treated with caution. As the reciprocal high similarity suggests a relation to Jmjd6, it might be possible to postulate an ancestral "Jmjd6" gene that encoded a larger protein with additional domains (like F-box domains), which then have been lost during evolution of Jmd6 genes in animalia. Alternatively, the evolution of Jmjd6 genes in animalia could have been an independent event and these genes are not related in any way to the proteins we identified in plants, fungi and protista. In the future it might be possible to consolidate on one hypothesis, given that more functional data about these hypothetical proteins are being gathered and higher quality sequence information becomes available for phylogenetic analysis.
Taken together, our integrated bioinformatic analysis provided new insights into possible structure-function relationships of Jmjd6 proteins and on the evolution of the whole superfamily of DSBH/JmjC domain containing proteins. Of note, the observations made on the protein model are not experimentally proven at the current state and await confirmation when the crystallographic structure of Jmjd6 has been obtained. In addition, we provide first evidence for a counteracting transcriptional regulative mechanism involving the co-localised Jmjd6 and 1110005A03Rik genes.
To investigate the splice variants of Jmjd6, the genomic sequence of Jmjd6 and adjacent genes on either side was extracted to perform an unrestricted e2g-based annotation of mouse ESTs with a seed-length of 18 and a minimum identity of 95% . The resulting list of ESTs that overlap with the known Jmjd6 coding-region was passed on to a self-developed script that automatically retrieves the sequences of all matches and performs an in-depth verification of each EST using Sim4 . The script sorts matches in different groups, according to the match quality. All matches with unidirectional splice sites, without internal gaps in the mRNA, and a <95% identity score to the corresponding genomic regions leaving aside the poly-A tail were used for manual annotation.
PipMaker sequence analysis  was done by extracting genomic sequences of 50 to 100 kb encompassing the Jmjd6 locus and neighbouring loci of mouse, human, bovine, armadillo, opossum, zebrafish, chicken, western clawed frog and puffer fish. Sequences were analysed by RepeatMasker  and positions of all coding sequences/exons in Jmjd6, 1110005A03Rik, and Sfrs2 were identified and annotated using the BLAST algorithm. The masked and annotated mouse sequence were then finally used for interspecies pairwise alignment with the sequences from the different species mentioned above using the program PipMaker .
To identify Jmjd6 orthologous sequences, a standard NCBI blastp-search against the nr database using amino acids 52–152 of the murine Jmjd6 protein was performed . Identified sequences were validated using the "reciprocal best BLAST hit" criterion in a Mus musculus-restricted blastp-search against the nr database .
SMART-parser, a Perl-script for batch queries against SMART , data retrieval and processing was developed to identify proteins with similar domains as Jmjd6. SMART and SMART-parser were used to automatically identify, extract and annotate the sequences of all JmjC-, Pfam-JmjC, Blast-JmjC-, P4Hc-, Pfam-2OG-FeII-Oxy- and Pfam-PhyH-domains for later processing. These domain names have been used in the different public databases to indicate the presence of the JmjC domain containing DSBH fold proteins.
Multiple sequence alignments were generated employing ClustalW with a default setup and five iterations of re-alignment. Visualisation, calculation of a background frequency corrected sequence logo, and calculation of amino acid conservation was done using LaTeX and TEXshade . A ClustalW alignment of all Jmjd6 orthologues (amino acids 44–329) identified in the prior approaches was used to generate a 3D-structure of Jmjd6 using the pipeline from the MPI Toolkit . To add 2OG and Fe(II) to the predicted structure and the predicted active site of Jmjd6, an iterative backbone annealing of the Hif1an structure (1h2k) on top of the Jmjd6 model was performed using the SwissPDB Viewer . Afterwards, the coordinates of 2OG and Fe(II) were added to the Jmjd6 model and energies were minimised using DeepView. 3D representations of molecules were rendered using PyMOL .
For a maximum likelihood phylogenetic analysis of Jmjd6 orthologues, we employed the PHYLIP toolbox . A ClustalW alignment of all proteins with significant homology to Jmjd6 was used as an input for ProML, using the Jones-Taylor-Thornton algorithm with an estimated constant rate of mutations, global rearrangements and the more accurate iterated branch lengths option. The primary protein sequences of Jmjd6/JMJD6 were analysed for sumoylation sites using a SUMOplot prediction with a cutt-off of 90% probability . Nuclear export signals (NES) were identified using the NetNES 1.1 server using a cut-off of 95% identity to the NES Hidden Markow Model in the database . Secondary structure predictions were generated using the Jpred web service .
Identification of Jmjd6 splice variants
RNA was isolated from different cells (HEK 293-T, A549, mouse embryonic fibroblasts (MEFs) and reverse transcribed. Murine splice variants were amplified using the forward primer 5'-GGTGGTGCCTCTTCCCAACAAA-3' and the reverse primers 5'-TCCAGGTCAGGGTTGGGACAC-3' and 5'-AGCTAGAAGAGTCGCTGGAGCTGTC-3'. The amplicons of the primer pairs had a product size of 363 bp (Mm3), 407 bp (Mm2), and 432 bp (Mm1), respectively. All fragments were analysed by sequencing and the sequences were submitted to GenBank (Acc. No. EF527404 and EF527405). Human cDNAs (C1244142, C1244152, C1244171-10, C1244035-10, C1244149, C1244264-10, C1244183, C1244246-10, C1244122-10, C1244260-10) were obtained from the BioChain Institute (Hayward, USA) and amplified according to the manufactures instructions using the forward primer 5'-GGGAACTCATCAAAGTGACCCGAGACG-3' and the reverse primers 5'-ACTGCCGCTGCCCGTGCTGTATC-3' and 5'-CGAGTCTGCGAGGACTGCCAACTCG-3'. The amplicons had a product size of 320 bp (Hs2), 344 bp (Hs1) and 438 bp (Hs3), respectively. All fragments were analysed by sequencing and the sequences were submitted to GenBank (Acc. No. EF527406 and EF527407).
Expression analysis of Jmjd6, 1110005A03Rik and Sfrs2
RNA was isolated and reverse transcribed. Expression of Jmjd6, 1110005A03Rik, Sfrs2 in embryos or cells were analysed using the primers 5'-GTT CCA GCT CGT CAG ACT CG-3' and 5'-TGC CCC TAA GAC ATG ACC AC-3' for Jmjd6, 5'-GCC ACC ACA AGA CAT CAT TCT TG-3' and 5'-TCA GAT AAT TCA GCT TTA TGC CAG G-3' for 1110005A03Rik, 5'-GCT CCA GAT CAA CCT CCA AG-3' and 5'-GCC ACC TGA GGC AGA TTA AA-3' for Sfrs2, respectively.
Analysis of the bidirectional transcriptional unit
An 841 bp and an 840 bp fragment was amplified from mouse C57BL/6J genomic DNA using the forward 5'-CTCGAGCTCTCGTAGTAGTTGTGCCGGGT-3' and reverse 5'-GCTAGCACTGTCCCTAAATGTGTCACTGGAGC-3' primer as well as the forward 3'-GCTAGCTCTCGTAGTAGTTGTGCCGGGT-5' and reverse 5'-CTCGAGACTGTCCCTAAATGTGTCACTGGAGC-3' primer, respectively. These fragments were integrated into a promoter-less luciferase expression vector (pGL3, Promega, Madison, USA) and these constructs were introduced into HEK 293-T cells, A549 cells or mouse embryonic fibroblasts (MEFs) using Nanofection (PAA, Pasching, Austria). After 24h of cultivation at 37°C and 5% CO2, cell extracts were generated to measure the luciferase-activity using the Dual-Luciferase-Reporter-Assay System (Promega) according to the manufactures instructions. The obtained results were normalized to pGL4.74 (Promega) luciferase-activity. The measurements were performed in triplicate and repeated at least three times.
We would like to thank Ivonne Wegener for excellent technical assistance. Dr. Richard Flavell (Yale University, New Haven, USA) and Dr. Yoshinori Fukui (Kyushu University, Fukuoka, Japan) for making available Jmjd6tm 1Flvknockout mice and RNA from Jmjd6tm 1Ysfkmutant embryos, respectively. We are thankful to Dr. Bastian Pasche (HZI, Braunschweig) for giving critical comments on the manuscript. This work was supported by a grant from Deutsche Forschungsgemeinschaft (DFG LE 2130/2-1) to AL.
- Takeuchi T, Yamazaki Y, Katoh-Fukui Y, Tsuchiya R, Kondo S, Motoyama J, Higashinakagawa T: Gene trap capture of a novel mouse gene, jumonji, required for neural tube formation. Genes Dev. 1995, 9: 1211-1222. 10.1101/gad.9.10.1211.PubMedView ArticleGoogle Scholar
- Balciunas D, Ronne H: Evidence of domain swapping within the jumonji family of transcription factors. Trends Biochem Sci. 2000, 25: 274-276. 10.1016/S0968-0004(00)01593-0.PubMedView ArticleGoogle Scholar
- Clissold PM, Ponting CP: JmjC: cupin metalloenzyme-like domains in jumonji, hairless and phospholipase A2beta. Trends Biochem Sci. 2001, 26: 7-9. 10.1016/S0968-0004(00)01700-X.PubMedView ArticleGoogle Scholar
- Klose RJ, Kallin EM, Zhang Y: JmjC-domain-containing proteins and histone demethylation. Nat Rev Genet. 2006, 7: 715-727. 10.1038/nrg1945.PubMedView ArticleGoogle Scholar
- Takeuchi T, Watanabe Y, Takano-Shimizu T, Kondo S: Roles of jumonji and jumonji family genes in chromatin regulation and development. Dev Dyn. 2006, 235: 2449-2459. 10.1002/dvdy.20851.PubMedView ArticleGoogle Scholar
- Clifton IJ, McDonough MA, Ehrismann D, Kershaw NJ, Granatino N, Schofield CJ: Structural studies on 2-oxoglutarate oxygenases and related double-stranded beta-helix fold proteins. J Inorg Biochem. 2006, 100: 644-669. 10.1016/j.jinorgbio.2006.01.024.PubMedView ArticleGoogle Scholar
- Dunwell JM, Culham A, Carter CE, Sosa-Aguirre CR, Goodenough PW: Evolution of functional diversity in the cupin superfamily. Trends Biochem Sci. 2001, 26: 740-746. 10.1016/S0968-0004(01)01981-8.PubMedView ArticleGoogle Scholar
- Schofield CJ, Zhang Z: Structural and mechanistic studies on 2-oxoglutarate-dependent oxygenases and related enzymes. Curr Opin Struct Biol. 1999, 9: 722-731. 10.1016/S0959-440X(99)00036-6.PubMedView ArticleGoogle Scholar
- Trewick SC, McLaughlin PJ, Allshire RC: Methylation: lost in hydroxylation?. EMBO Rep. 2005, 6: 315-320. 10.1038/sj.embor.7400379.PubMedView ArticleGoogle Scholar
- Lee Y, Song AJ, Baker R, Micales B, Conway SJ, Lyons GE: Jumonji, a nuclear protein that is necessary for normal heart development. Circ Res. 2000, 86: 932-938.PubMedView ArticleGoogle Scholar
- Takeuchi T, Kojima M, Nakajima K, Kondo S: jumonji gene is essential for the neurulation and cardiac development of mouse embryos with a C3H/He background. Mech Dev. 1999, 86: 29-38. 10.1016/S0925-4773(99)00100-8.PubMedView ArticleGoogle Scholar
- Toyoda M, Kojima M, Takeuchi T: Jumonji is a nuclear protein that participates in the negative regulation of cell growth. Biochem Biophys Res Commun. 2000, 274: 332-336. 10.1006/bbrc.2000.3138.PubMedView ArticleGoogle Scholar
- Benevolenskaya EV, Murray HL, Branton P, Young RA, Kaelin WG: Binding of pRB to the PHD protein RBP2 promotes cellular differentiation. Mol Cell. 2005, 18: 623-635. 10.1016/j.molcel.2005.05.012.PubMedView ArticleGoogle Scholar
- Hsieh JC, Sisk JM, Jurutka PW, Haussler CA, Slater SA, Haussler MR, Thompson CC: Physical and functional interaction between the vitamin D receptor and hairless corepressor, two proteins required for hair cycling. J Biol Chem. 2003, 278: 38665-38674. 10.1074/jbc.M304886200.PubMedView ArticleGoogle Scholar
- Moraitis AN, Giguere V, Thompson CC: Novel mechanism of nuclear receptor corepressor interaction dictated by activation function 2 helix determinants. Mol Cell Biol. 2002, 22: 6831-6841. 10.1128/MCB.22.19.6831-6841.2002.PubMedView ArticleGoogle Scholar
- Potter GB, Beaudoin GM, DeRenzo CL, Zarach JM, Chen SH, Thompson CC: The hairless gene mutated in congenital hair loss disorders encodes a novel nuclear receptor corepressor. Genes Dev. 2001, 15: 2687-2701. 10.1101/gad.916701.PubMedView ArticleGoogle Scholar
- Ahmad W, Faiyaz ul Haque M, Brancolini V, Tsou HC, ul Haque S, Lam H, Aita VM, Owen J, deBlaquiere M, Frank J: Alopecia universalis associated with a mutation in the human hairless gene. Science. 1998, 279: 720-724. 10.1126/science.279.5351.720.PubMedView ArticleGoogle Scholar
- Cichon S, Anker M, Vogt IR, Rohleder H, Putzstuck M, Hillmer A, Farooq SA, Al-Dhafri KS, Ahmad M, Haque S: Cloning, genomic organization, alternative transcripts and mutational analysis of the gene responsible for autosomal recessive universal congenital alopecia. Hum Mol Genet. 1998, 7: 1671-1679. 10.1093/hmg/7.11.1671.PubMedView ArticleGoogle Scholar
- Zarach JM, Beaudoin GM, Coulombe PA, Thompson CC: The co-repressor hairless has a role in epithelial cell differentiation in the skin. Development. 2004, 131: 4189-4200. 10.1242/dev.01303.PubMedView ArticleGoogle Scholar
- Martin C, Zhang Y: The diverse functions of histone lysine methylation. Nat Rev Mol Cell Biol. 2005, 6: 838-849. 10.1038/nrm1761.PubMedView ArticleGoogle Scholar
- Tsukada Y, Fang J, Erdjument-Bromage H, Warren ME, Borchers CH, Tempst P, Zhang Y: Histone demethylation by a family of JmjC domain-containing proteins. Nature. 2006, 439: 811-816. 10.1038/nature04433.PubMedView ArticleGoogle Scholar
- Chen Z, Zang J, Whetstine J, Hong X, Davrazou F, Kutateladze TG, Simpson M, Mao Q, Pan CH, Dai S: Structural insights into histone demethylation by JMJD2 family members. Cell. 2006, 125: 691-702. 10.1016/j.cell.2006.04.024.PubMedView ArticleGoogle Scholar
- Whetstine JR, Nottke A, Lan F, Huarte M, Smolikov S, Chen Z, Spooner E, Li E, Zhang G, Colaiacovo M, Shi Y: Reversal of histone lysine trimethylation by the JMJD2 family of histone demethylases. Cell. 2006, 125: 467-481. 10.1016/j.cell.2006.03.028.PubMedView ArticleGoogle Scholar
- Klose RJ, Gardner KE, Liang G, Erdjument-Bromage H, Tempst P, Zhang Y: Demethylation of histone H3K36 and H3K9 by Rph1: a vestige of an H3K9 methylation system in Saccharomyces cerevisiae?. Mol Cell Biol. 2007, 27: 3951-3961. 10.1128/MCB.02180-06.PubMedView ArticleGoogle Scholar
- Tu S, Bulloch EM, Yang L, Ren C, Huang WC, Hsu PH, Chen CH, Liao CL, Yu HM, Lo WS: Identification of histone demethylases in Sacchromyces cerevisiae. J Biol Chem. 2007, 282: 14262-14271. 10.1074/jbc.M609900200.PubMedView ArticleGoogle Scholar
- Yamane K, Tateishi K, Klose RJ, Fang J, Fabrizio LA, Erdjument-Bromage H, Taylor-Papadimitriou J, Tempst P, Zhang Y: PLU-1 is an H3K4 demethylase involved in transcriptional repression and breast cancer cell proliferation. Mol Cell. 2007, 25: 801-812. 10.1016/j.molcel.2007.03.001.PubMedView ArticleGoogle Scholar
- Cloos PA, Christensen J, Agger K, Maiolica A, Rappsilber J, Antal T, Hansen KH, Helin K: The putative oncogene GASC1 demethylates tri- and dimethylated lysine 9 on histone H3. Nature. 2006, 442: 307-311. 10.1038/nature04837.PubMedView ArticleGoogle Scholar
- Fodor BD, Kubicek S, Yonezawa M, O'Sullivan RJ, Sengupta R, Perez-Burgos L, Opravil S, Mechtler K, Schotta G, Jenuwein T: Jmjd2b antagonizes H3K9 trimethylation at pericentric heterochromatin in mammalian cells. Genes Dev. 2006, 20: 1557-1562. 10.1101/gad.388206.PubMedView ArticleGoogle Scholar
- Klose RJ, Yamane K, Bae Y, Zhang D, Erdjument-Bromage H, Tempst P, Wong J, Zhang Y: The transcriptional repressor JHDM3A demethylates trimethyl histone H3 lysine 9 and lysine 36. Nature. 2006, 442: 312-316. 10.1038/nature04853.PubMedView ArticleGoogle Scholar
- Wissmann M, Yin N, Muller JM, Greschik H, Fodor BD, Jenuwein T, Vogler C, Schneider R, Gunther T, Buettner R: Cooperative demethylation by JMJD2C and LSD1 promotes androgen receptor-dependent gene expression. Nat Cell Biol. 2007, 9: 347-353. 10.1038/ncb1546.PubMedView ArticleGoogle Scholar
- Christensen J, Agger K, Cloos PA, Pasini D, Rose S, Sennels L, Rappsilber J, Hansen KH, Salcini AE, Helin K: RBP2 belongs to a family of demethylases, specific for tri-and dimethylated lysine 4 on histone 3. Cell. 2007, 128: 1063-1076. 10.1016/j.cell.2007.02.003.PubMedView ArticleGoogle Scholar
- Klose RJ, Yan Q, Tothova Z, Yamane K, Erdjument-Bromage H, Tempst P, Gilliland DG, Zhang Y, Kaelin WG: The retinoblastoma binding protein RBP2 is an H3K4 demethylase. Cell. 2007, 128: 889-900. 10.1016/j.cell.2007.02.013.PubMedView ArticleGoogle Scholar
- Iwase S, Lan F, Bayliss P, de la Torre-Ubieta L, Huarte M, Qi HH, Whetstine JR, Bonni A, Roberts TM, Shi Y: The X-linked mental retardation gene SMCX/JARID1C defines a family of histone H3 lysine 4 demethylases. Cell. 2007, 128: 1077-1088. 10.1016/j.cell.2007.02.017.PubMedView ArticleGoogle Scholar
- Lee MG, Norman J, Shilatifard A, Shiekhattar R: Physical and functional association of a trimethyl H3K4 demethylase and Ring6a/MBLR, a polycomb-like protein. Cell. 2007, 128: 877-887. 10.1016/j.cell.2007.02.004.PubMedView ArticleGoogle Scholar
- Liang G, Klose RJ, Gardner KE, Zhang Y: Yeast Jhd2p is a histone H3 Lys4 trimethyl demethylase. Nat Struct Mol Biol. 2007, 14: 243-245. 10.1038/nsmb1204.PubMedView ArticleGoogle Scholar
- Seward DJ, Cubberley G, Kim S, Schonewald M, Zhang L, Tripet B, Bentley DL: Demethylation of trimethylated histone H3 Lys4 in vivo by JARID1 JmjC proteins. Nat Struct Mol Biol. 2007, 14: 240-242. 10.1038/nsmb1200.PubMedView ArticleGoogle Scholar
- Eissenberg JC, Lee MG, Schneider J, Ilvarsonn A, Shiekhattar R, Shilatifard A: The trithorax-group gene in Drosophila little imaginal discs encodes a trimethylated histone H3 Lys4 demethylase. Nat Struct Mol Biol. 2007, 14: 344-346. 10.1038/nsmb1217.PubMedView ArticleGoogle Scholar
- Lee N, Zhang J, Klose RJ, Erdjument-Bromage H, Tempst P, Jones RS, Zhang Y: The trithorax-group protein Lid is a histone H3 trimethyl-Lys4 demethylase. Nat Struct Mol Biol. 2007, 14: 341-343. 10.1038/nsmb1216.PubMedView ArticleGoogle Scholar
- Secombe J, Li L, Carlos L, Eisenman RN: The Trithorax group protein Lid is a trimethyl histone H3K4 demethylase required for dMyc-induced cell growth. Genes Dev. 2007, 21: 537-551. 10.1101/gad.1523007.PubMedView ArticleGoogle Scholar
- Huarte M, Lan F, Kim T, Vaughn MW, Zaratiegui M, Martienssen RA, Buratowski S, Shi Y: The fission yeast JMJ2 reverses histone H3 lysine 4 tri-methylation. J Biol Chem. 2007, 282: 21662-21670. 10.1074/jbc.M703897200.PubMedView ArticleGoogle Scholar
- Agger K, Cloos PA, Christensen J, Pasini D, Rose S, Rappsilber J, Issaeva I, Canaani E, Salcini AE, Helin K: UTX and JMJD3 are histone H3K27 demethylases involved in HOX gene regulation and development. Nature. 2007, 449: 731-734. 10.1038/nature06145.PubMedView ArticleGoogle Scholar
- De Santa F, Totaro MG, Prosperini E, Notarbartolo S, Testa G, Natoli G: The histone H3 lysine-27 demethylase Jmjd3 links inflammation to inhibition of polycomb-mediated gene silencing. Cell. 2007, 130: 1083-1094. 10.1016/j.cell.2007.08.019.PubMedView ArticleGoogle Scholar
- Lan F, Bayliss PE, Rinn JL, Whetstine JR, Wang JK, Chen S, Iwase S, Alpatov R, Issaeva I, Canaani E: A histone H3 lysine 27 demethylase regulates animal posterior development. Nature. 2007, 449: 689-694. 10.1038/nature06192.PubMedView ArticleGoogle Scholar
- Lee MG, Villa R, Trojer P, Norman J, Yan KP, Reinberg D, Di Croce L, Shiekhattar R: Demethylation of H3K27 regulates polycomb recruitment and H2A ubiquitination. Science. 2007, 318: 447-450. 10.1126/science.1149042.PubMedView ArticleGoogle Scholar
- Yamane K, Toumazou C, Tsukada Y, Erdjument-Bromage H, Tempst P, Wong J, Zhang Y: JHDM2A, a JmjC-containing H3K9 demethylase, facilitates transcription activation by androgen receptor. Cell. 2006, 125: 483-495. 10.1016/j.cell.2006.03.027.PubMedView ArticleGoogle Scholar
- Zofall M, Grewal SI: Swi6/HP1 recruits a JmjC domain protein to facilitate transcription of heterochromatic repeats. Mol Cell. 2006, 22: 681-692. 10.1016/j.molcel.2006.05.010.PubMedView ArticleGoogle Scholar
- Huang Y, Fang J, Bedford MT, Zhang Y, Xu RM: Recognition of histone H3 lysine-4 methylation by the double tudor domain of JMJD2A. Science. 2006, 312: 748-751. 10.1126/science.1125162.PubMedView ArticleGoogle Scholar
- Dann CE, Bruick RK, Deisenhofer J: Structure of factor-inhibiting hypoxia-inducible factor 1: An asparaginyl hydroxylase involved in the hypoxic response pathway. Proc Natl Acad Sci USA. 2002, 99: 15351-15356. 10.1073/pnas.202614999.PubMedView ArticleGoogle Scholar
- Elkins JM, Hewitson KS, McNeill LA, Seibel JF, Schlemminger I, Pugh CW, Ratcliffe PJ, Schofield CJ: Structure of factor-inhibiting hypoxia-inducible factor (HIF) reveals mechanism of oxidative modification of HIF-1 alpha. J Biol Chem. 2003, 278: 1802-1806. 10.1074/jbc.C200644200.PubMedView ArticleGoogle Scholar
- Schofield CJ, Ratcliffe PJ: Oxygen sensing by HIF hydroxylases. Nat Rev Mol Cell Biol. 2004, 5: 343-354. 10.1038/nrm1366.PubMedView ArticleGoogle Scholar
- Falnes PO, Johansen RF, Seeberg E: AlkB-mediated oxidative demethylation reverses DNA damage in Escherichia coli. Nature. 2002, 419: 178-182. 10.1038/nature01048.PubMedView ArticleGoogle Scholar
- Trewick SC, Henshaw TF, Hausinger RP, Lindahl T, Sedgwick B: Oxidative demethylation by Escherichia coli AlkB directly reverts DNA base damage. Nature. 2002, 419: 174-178. 10.1038/nature00908.PubMedView ArticleGoogle Scholar
- Aravind L, Koonin EV: The DNA-repair protein AlkB, EGL-9, and leprecan define new families of 2-oxoglutarate- and iron-dependent dioxygenases. Genome Biol. 2001, 2: RESEARCH0007-10.1186/gb-2001-2-3-research0007.PubMedView ArticleGoogle Scholar
- Böse J, Gruber AD, Helming L, Schiebe S, Wegener I, Hafner M, Beales M, Kontgen F, Lengeling A: The phosphatidylserine receptor has essential functions during embryogenesis but not in apoptotic cell removal. J Biol. 2004, 3: 15-10.1186/jbiol10.PubMedView ArticleGoogle Scholar
- Schneider JE, Böse J, Bamforth SD, Gruber AD, Broadbent C, Clarke K, Neubauer S, Lengeling A, Bhattacharya S: Identification of cardiac malformations in mice lacking Ptdsr using a novel high-throughput magnetic resonance imaging technique. BMC Dev Biol. 2004, 4: 16-10.1186/1471-213X-4-16.PubMedView ArticleGoogle Scholar
- Fadok VA, Bratton DL, Rose DM, Pearson A, Ezekewitz RA, Henson PM: A receptor for phosphatidylserine-specific clearance of apoptotic cells. Nature. 2000, 405: 85-90. 10.1038/35011084.PubMedView ArticleGoogle Scholar
- Cikala M, Alexandrova O, David CN, Proschel M, Stiening B, Cramer P, Böttger A: The phosphatidylserine receptor from Hydra is a nuclear protein with potential Fe(II) dependent oxygenase activity. BMC Cell Biol. 2004, 5: 26-10.1186/1471-2121-5-26.PubMedView ArticleGoogle Scholar
- Cui P, Qin B, Liu N, Pan G, Pei D: Nuclear localization of the phosphatidylserine receptor protein via multiple nuclear localization signals. Exp Cell Res. 2004, 293: 154-163. 10.1016/j.yexcr.2003.09.023.PubMedView ArticleGoogle Scholar
- Krieser RJ, Moore FE, Dresnek D, Pellock BJ, Patel R, Huang A, Brachmann C, White K: The Drosophila homolog of the putative phosphatidylserine receptor functions to inhibit apoptosis. Development. 2007, 134: 2407-2414. 10.1242/dev.02860.PubMedView ArticleGoogle Scholar
- Mitchell JE, Cvetanovic M, Tibrewal N, Patel V, Colamonici OR, Li MO, Flavell RA, Levine JS, Birge RB, Ucker DS: The presumptive phosphatidylserine receptor is dispensable for innate anti-inflammatory recognition and clearance of apoptotic cells. J Biol Chem. 2006, 281: 5718-5725. 10.1074/jbc.M509775200.PubMedView ArticleGoogle Scholar
- Tibrewal N, Liu T, Li H, Birge RB: Characterization of the biochemical and biophysical properties of the phosphatidylserine receptor (PS-R) gene product. Mol Cell Biochem. 2007, 304: 119-125. 10.1007/s11010-007-9492-8.PubMedView ArticleGoogle Scholar
- Li MO, Sarkisian MR, Mehal WZ, Rakic P, Flavell RA: Phosphatidylserine receptor is required for clearance of apoptotic cells. Science. 2003, 302: 1560-1563. 10.1126/science.1087621.PubMedView ArticleGoogle Scholar
- Kunisaki Y, Masuko S, Noda M, Inayoshi A, Sanui T, Harada M, Sasazuki T, Fukui Y: Defective fetal liver erythropoiesis and T lymphopoiesis in mice lacking the phosphatidylserine receptor. Blood. 2004, 103: 3362-3364. 10.1182/blood-2003-09-3245.PubMedView ArticleGoogle Scholar
- Fackelmayer FO: Protein arginine methyltransferases: guardians of the Arg?. Trends Biochem Sci. 2005, 30: 666-671.PubMedView ArticleGoogle Scholar
- Bermingham JR, Arden KC, Naumova AK, Sapienza C, Viars CS, Fu XD, Khotz J, Manley JL, Rosenfeld MG: Chromosomal localization of mouse and human genes encoding the splicing factors ASF/SF2 (SFRS1) and SC-35 (SFRS2). Genomics. 1995, 29: 70-79. 10.1006/geno.1995.1216.PubMedView ArticleGoogle Scholar
- Schwartz S, Zhang Z, Frazer KA, Smit A, Riemer C, Bouck J, Gibbs R, Hardison R, Miller W: PipMaker – a web server for aligning two genomic DNA sequences. Genome Res. 2000, 10: 577-586. 10.1101/gr.10.4.577.PubMedView ArticleGoogle Scholar
- Krüger J, Sczyrba A, Kurtz S, Giegerich R: e2g: an interactive web-based server for efficiently mapping large EST and cDNA sets to genomic sequences. Nucleic Acids Res. 2004, 32: W301-304. 10.1093/nar/gkh478.PubMedView ArticleGoogle Scholar
- Florea L, Hartzell G, Zhang Z, Rubin GM, Miller W: A computer program for aligning a cDNA sequence with a genomic DNA sequence. Genome Res. 1998, 8: 967-974.PubMedGoogle Scholar
- Fadok VA, Xue D, Henson P: If phosphatidylserine is the death knell, a new phosphatidylserine-specific receptor is the bellringer. Cell Death Differ. 2001, 8: 582-587. 10.1038/sj.cdd.4400856.PubMedView ArticleGoogle Scholar
- Huntley MA, Golding GB: Selection and slippage creating serine homopolymers. Mol Biol Evol. 2006, 23: 2017-2025. 10.1093/molbev/msl073.PubMedView ArticleGoogle Scholar
- Amrani N, Sachs MS, Jacobson A: Early nonsense: mRNA decay solves a translational problem. Nat Rev Mol Cell Biol. 2006, 7: 415-425. 10.1038/nrm1942.PubMedView ArticleGoogle Scholar
- Tatusov RL, Koonin EV, Lipman DJ: A genomic perspective on protein families. Science. 1997, 278: 631-637. 10.1126/science.278.5338.631.PubMedView ArticleGoogle Scholar
- Hubbard TJ, Aken BL, Beal K, Ballester B, Caccamo M, Chen Y, Clarke L, Coates G, Cunningham F, Cutts T: Ensembl 2007. Nucleic Acids Res. 2007, 35: D610-617. 10.1093/nar/gkl996.PubMedView ArticleGoogle Scholar
- Söding J: Protein homology detection by HMM-HMM comparison. Bioinformatics. 2005, 21: 951-960. 10.1093/bioinformatics/bti125.PubMedView ArticleGoogle Scholar
- Sali A, Blundell TL: Comparative protein modelling by satisfaction of spatial restraints. J Mol Biol. 1993, 234: 779-815. 10.1006/jmbi.1993.1626.PubMedView ArticleGoogle Scholar
- Britsch L, Dedio J, Saedler H, Forkmann G: Molecular characterization of flavanone 3 beta-hydroxylases. Consensus sequence, comparison with related enzymes and the role of conserved histidine residues. Eur J Biochem. 1993, 217: 745-754. 10.1111/j.1432-1033.1993.tb18301.x.PubMedView ArticleGoogle Scholar
- Hogan DA, Smith SR, Saari EA, McCracken J, Hausinger RP: Site-directed mutagenesis of 2,4-dichlorophenoxyacetic acid/alpha-ketoglutarate dioxygenase. Identification of residues involved in metallocenter formation and substrate binding. J Biol Chem. 2000, 275: 12400-12409. 10.1074/jbc.275.17.12400.PubMedView ArticleGoogle Scholar
- Lee DH, Jin SG, Cai S, Chen Y, Pfeifer GP, O'Connor TR: Repair of methylation damage in DNA and RNA by mammalian AlkB homologues. J Biol Chem. 2005, 280: 39448-39459. 10.1074/jbc.M509881200.PubMedView ArticleGoogle Scholar
- Yu B, Edstrom WC, Benach J, Hamuro Y, Weber PC, Gibney BR, Hunt JF: Crystal structures of catalytic complexes of the oxidative DNA/RNA repair enzyme AlkB. Nature. 2006, 439: 879-884. 10.1038/nature04561.PubMedView ArticleGoogle Scholar
- Klose RJ, Zhang Y: Regulation of histone methylation by demethylimination and demethylation. Nat Rev Mol Cell Biol. 2007, 8: 307-318. 10.1038/nrm2143.PubMedView ArticleGoogle Scholar
- Ozer A, Bruick RK: Non-heme dioxygenases: cellular sensors and regulators jelly rolled into one?. Nat Chem Biol. 2007, 3: 144-153. 10.1038/nchembio863.PubMedView ArticleGoogle Scholar
- Weitzman JB: The curious world of apoptotic cell clearance. J Biol. 2004, 3: 13-10.1186/jbiol13.View ArticleGoogle Scholar
- Williamson P, Schlegel RA: Hide and seek: the secret identity of the phosphatidylserine receptor. J Biol. 2004, 3: 14-10.1186/jbiol14.PubMedView ArticleGoogle Scholar
- Wolf A, Schmitz C, Bottger A: Changing story of the receptor for phosphatidylserine-dependent clearance of apoptotic cells. EMBO Rep. 2007, 8: 465-469. 10.1038/sj.embor.7400956.PubMedView ArticleGoogle Scholar
- Aravind L, Landsman D: AT-hook motifs identified in a wide variety of DNA-binding proteins. Nucleic Acids Res. 1998, 26: 4413-4421. 10.1093/nar/26.19.4413.PubMedView ArticleGoogle Scholar
- Chang B, Chen Y, Zhao Y, Bruick RK: JMJD6 is a histone arginine demethylase. Science. 2007, 318: 444-447. 10.1126/science.1145801.PubMedView ArticleGoogle Scholar
- Guccione E, Bassi C, Casadio F, Martinato F, Cesaroni M, Schuchlautz H, Luscher B, Amati B: Methylation of histone H3R2 by PRMT6 and H3K4 by an MLL complex are mutually exclusive. Nature. 2007, 449: 933-937. 10.1038/nature06166.PubMedView ArticleGoogle Scholar
- Kirmizis A, Santos-Rosa H, Penkett CJ, Singer MA, Vermeulen M, Mann M, Bahler J, Green RD, Kouzarides T: Arginine methylation at histone H3R2 controls deposition of H3K4 trimethylation. Nature. 2007, 449: 928-932. 10.1038/nature06160.PubMedView ArticleGoogle Scholar
- Ng SS, Kavanagh KL, McDonough MA, Butler D, Pilka ES, Lienard BM, Bray JE, Savitsky P, Gileadi O, von Delft F: Crystal structures of histone demethylase JMJD2A reveal basis for substrate specificity. Nature. 2007, 448: 87-91. 10.1038/nature05971.PubMedView ArticleGoogle Scholar
- Kim J, Daniel J, Espejo A, Lake A, Krishna M, Xia L, Zhang Y, Bedford MT: Tudor, MBT and chromo domains gauge the degree of lysine methylation. EMBO Rep. 2006, 7: 397-403.PubMedGoogle Scholar
- Li H, Ilin S, Wang W, Duncan EM, Wysocka J, Allis CD, Patel DJ: Molecular basis for site-specific read-out of histone H3K4me3 by the BPTF PHD finger of NURF. Nature. 2006, 442: 91-95. 10.1038/nature05020.PubMedView ArticleGoogle Scholar
- Pena PV, Davrazou F, Shi X, Walter KL, Verkhusha VV, Gozani O, Zhao R, Kutateladze TG: Molecular mechanism of histone H3K4me3 recognition by plant homeodomain of ING2. Nature. 2006, 442: 100-103.PubMedGoogle Scholar
- Shi X, Hong T, Walter KL, Ewalt M, Michishita E, Hung T, Carney D, Pena P, Lan F, Kaadige MR: ING2 PHD domain links histone H3 lysine 4 methylation to active gene repression. Nature. 2006, 442: 96-99. 10.1038/nature05140.PubMedView ArticleGoogle Scholar
- Wysocka J, Swigut T, Xiao H, Milne TA, Kwon SY, Landry J, Kauer M, Tackett AJ, Chait BT, Badenhorst P: A PHD finger of NURF couples histone H3 lysine 4 trimethylation with chromatin remodelling. Nature. 2006, 442: 86-90.PubMedGoogle Scholar
- Hong JR, Lin GH, Lin CJ, Wang WP, Lee CC, Lin TL, Wu JL: Phosphatidylserine receptor is required for the engulfment of dead apoptotic cells and for normal embryonic development in zebrafish. Development. 2004, 131: 5417-5427. 10.1242/dev.01409.PubMedView ArticleGoogle Scholar
- Wang X, Wu YC, Fadok VA, Lee MC, Gengyo-Ando K, Cheng LC, Ledwich D, Hsu PK, Chen JY, Chou BK: Cell corpse engulfment mediated by C. elegans phosphatidylserine receptor through CED-5 and CED-12. Science. 2003, 302: 1563-1566. 10.1126/science.1087641.PubMedView ArticleGoogle Scholar
- RepeatMasker. [http://www.repeatmasker.org]
- NCBI-BLAST. [http://www.ncbi.nlm.nih.gov/BLAST]
- SMART. [http://smart.embl-heidelberg.de]
- Beitz E: TEXshade: shading and labeling of multiple sequence alignments using LATEX2 epsilon. Bioinformatics. 2000, 16: 135-139. 10.1093/bioinformatics/16.2.135.PubMedView ArticleGoogle Scholar
- Biegert A, Mayer C, Remmert M, Soding J, Lupas AN: The MPI Bioinformatics Toolkit for protein sequence analysis. Nucleic Acids Res. 2006, 34: W335-339. 10.1093/nar/gkl217.PubMedView ArticleGoogle Scholar
- Swiss-Pdb-Viewer. [http://expasy.org/spdbv]
- PyMOL. [http://pymol.sourceforge.net/]
- PHYLIP. [http://evolution.genetics.washington.edu/phylip.html]
- SUMOplot. [http://www.abgent.com.cn/doc/sumoplot/login.asp]
- NetNES. [http://www.cbs.dtu.dk/services/NetNES]
- Jpred-3. [http://www.compbio.dundee.ac.uk/~www-jpred]
- NCBI-Taxonomy. [http://www.ncbi.nlm.nih.gov/Taxonomy/]
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.