Computational identification of novel biochemical systems involved in oxidation,glycosylation and other complex modifications of bases in DNA

Discovery of the TET/JBP family of dioxygenases that modify bases in DNA has sparked considerable interest in novel DNA base modifications and their biological roles. Using sensitive sequence and structure analyses combined with contextual information from comparative genomics, we computationally characterize over 12 novel biochemical systems for DNA modifications. We predict previously unidentified enzymes, such as the kinetoplastid J-base generating glycosyltransferase (and its homolog GREB1), the catalytic specificity of bacteriophage TET/JBP proteins and their role in complex DNA base modifications. We also predict the enzymes involved in synthesis of hypermodified bases such as alpha-glutamylthymine and alpha-putrescinylthymine that have remained enigmatic for several decades. Moreover, the current analysis suggests that bacteriophages and certain nucleo-cytoplasmic large DNA viruses contain an unexpectedly diverse range of DNA modification systems, in addition to those using previously characterized enzymes such as Dam, Dcm, TET/JBP, pyrimidine hydroxymethylases, Mom and glycosyltransferases. These include enzymes generating modified bases such as deazaguanines related to queuine and archaeosine, pyrimidines comparable with lysidine, those derived using modified S-adenosyl methionine derivatives and those using TET/JBP-generated hydroxymethyl pyrimidines as biosynthetic starting points. We present evidence that some of these modification systems are also widely dispersed across prokaryotes and certain eukaryotes such as basidiomycetes, chlorophyte and stramenopile alga, where they could serve as novel epigenetic marks for regulation or discrimination of self from non-self DNA. Our study extends the role of the PUA-like fold domains in recognition of modified nucleic acids and predicts versions of the ASCH and EVE domains to be novel ‘readers’ of modified bases in DNA. These results open opportunities for the investigation of the biology of these systems and their use in biotechnology.     

The protein that binds to DNA base J in trypanosomatids has features of a thymidine hydroxylase

Trypanosomatids contain an unusual DNA base J (beta-d-glucosylhydroxymethyluracil), which replaces a fraction of thymine in telomeric and other DNA repeats. To determine the function of base J, we have searched for enzymes that catalyze J biosynthesis. We present evidence that a protein that binds to J in DNA, the J-binding protein 1 (JBP1), may also catalyze the first step in J biosynthesis, the conversion of thymine in DNA into hydroxymethyluracil. We show that JBP1 belongs to the family of Fe(2+) and 2-oxoglutarate-dependent dioxygenases and that replacement of conserved residues putatively involved in Fe(2+) and 2-oxoglutarate-binding inactivates the ability of JBP1 to contribute to J synthesis without affecting its ability to bind to J-DNA. We propose that JBP1 is a thymidine hydroxylase responsible for the local amplification of J inserted by JBP2, another putative thymidine hydroxylase.     

J-binding protein increases the level and retention of the unusual base J in trypanosome DNA

The nuclear DNA of Trypanosoma brucei and other kinetoplastid flagellates contains the unusual base beta-d-glucosyl-hydroxymethyluracil, called J, replacing part of the thymine in repetitive sequences. We have described a 100 kDa protein that specifically binds to J in duplex DNA. We have now disrupted the genes for this J-binding protein (JBP) in T. brucei. The disruption does not affect growth, gene expression or the stability of some repetitive DNA sequences. Unexpectedly, however, the JBP KO trypanosomes contain only about 5% of the wild-type level of J in their DNA. Excess J, randomly introduced into T. brucei DNA by growing the cells in the presence of the J precursor 5-hydroxymethyldeoxyuridine, is lost by simple dilution as the KO trypanosomes multiply, showing that JBP does not protect J against removal. In contrast, cells containing JBP lose excess J only sluggishly. We conclude that JBP is able to activate the thymine modification enzymes to introduce additional J in regions of DNA already containing a basal level of J. We propose that JBP is a novel DNA modification maintenance protein.     

Ascorbate-induced generation of 5-hydroxymethylcytosine is unaffected by varying levels of iron and 2-oxoglutarate

Tet (ten–eleven translocation) methylcytosine dioxygenases, which belong to the iron and 2-oxoglutarate (2OG)-dependent dioxygenase superfamily, convert 5-methylcytosine (5mC) to 5-hydroxymethylcytosine (5hmC) in DNA. We recently reported that ascorbate (vitamin C) induces Tet-mediated generation of 5hmC. To initially delineate the role of ascorbate on 5hmC generation, we analyzed whether the effect of ascorbate is dependent upon the conditions of other components involved in the hydroxylation of 5mC catalyzed by Tet. We found that removing iron from the culture medium did not affect the induction of 5hmC by ascorbate (10 μM) in mouse embryonic fibroblasts (MEFs). The effect of ascorbate did not involve an increased expression of Tet1–3 or isocitrate dehydrogenases (IDH1–2), the enzymes responsible for producing 2OG. Interestingly, MEFs cultured with different concentrations of glucose, a major precursor of 2OG, exhibited nearly identical responses to ascorbate treatment. Further, blocking the uptake of the reduced form of vitamin C, ascorbic acid, through the sodium-dependent vitamin C transporters (SVCTs) inhibited the effect of ascorbate on 5hmC. However, inhibition of the facilitative glucose transporters (GLUTs), which mediate the incorporation of the oxidized form of vitamin C, dehydroascorbic acid (DHA), did not modify the ability of ascorbate to induce 5hmC generation. These results indicate that the effect of ascorbate on 5hmC is not dependent upon iron uptake, the expression of Tet and IDH, or the production of 2OG, suggesting that ascorbate may directly participate in the generation of 5hmC, most likely as a cofactor of Tet.     

The Schizosaccharomyces pombe AlkB homolog Abh1 exhibits AP lyase activity but no demethylase activity

2-Oxoglutarate (2OG) and iron (Fe(II)) dependent dioxygenases catalyze a wide range of biological oxidations, including hydroxylation and demethylation of proteins and nucleic acids. AlkB from Escherichia coli directly reverses certain methyl lesions in DNA, and defines a subfamily of 2OG/Fe(II) dioxygenases that has so far been shown to be involved in both nucleic acid repair and modification. The human genome encodes nine AlkB homologs and the function of most of these is still unknown. The fission yeast Schizosaccharomyces pombe has two AlkB homologs and here we have addressed the function of one of these, Abh1, which appears not to possess a classical AlkB-like repair activity. No enzymatic activity was found toward methylated DNA or etheno adducts, nor was the yeast abh1- mutant sensitive toward alkylating agents. Interestingly, heterologous expression of E. coli AlkB protected the fission yeast cells from alkylation induced cytotoxicity, suggesting that S. pombe lacks systems for efficient repair of lesions that are AlkB substrates. Further, we show that Abh1 possesses an unexpected DNA incision activity at apurinic/apyrimidinic (AP) sites. This AP lyase activity did not depend on 2OG and Fe(II) and was not repressed by dioxygenase inhibitors. Survival and complementation analyses failed to reveal any biological role for AP lyase cleavage by Abh1. It appears that in vitro AP lyase activity can be detected for a number of enzymes belonging to structurally and functionally unrelated families, but the in vivo significance of such activities may be questionable.     

Bioinformatics and functional analysis define four distinct groups of AlkB DNA-dioxygenases in bacteria

The iron(II)- and 2-oxoglutarate (2OG)-dependent dioxygenase AlkB from Escherichia coli (EcAlkB) repairs alkylation damage in DNA by direct reversal. EcAlkB substrates include methylated bases, such as 1-methyladenine (m¹A) and 3-methylcytosine (m³C), as well as certain bulkier lesions, for example the exocyclic adduct 1,N⁶-ethenoadenine (εA). EcAlkB is the only bacterial AlkB protein characterized to date, and we here present an extensive bioinformatics and functional analysis of bacterial AlkB proteins. Based on sequence phylogeny, we show that these proteins can be subdivided into four groups: denoted 1A, 1B, 2A and 2B; each characterized by the presence of specific conserved amino acid residues in the putative nucleotide-recognizing domain. A scattered distribution of AlkB proteins from the four different groups across the bacterial kingdom indicates a substantial degree of horizontal transfer of AlkB genes. DNA repair activity was associated with all tested recombinant AlkB proteins. Notably, both a group 2B protein from Xanthomonas campestris and a group 2A protein from Rhizobium etli repaired etheno adducts, but had negligible activity on methylated bases. Our data indicate that the majority, if not all, of the bacterial AlkB proteins are DNA repair enzymes, and that some of these proteins do not primarily target methylated bases.     

Non-random association of transposable elements with duplicated genomic blocks in Arabidopsis thaliana

The genome of Arabidopsis thaliana is known to contain numerous open reading frames apparently encoding transposases. In order to test the hypothesis that transposable elements have played a role in segmental duplication in this species, we compared the distribution of transposable elements with that of genomic windows that shared gene families to a greater extent than expected by chance. Phylogenetic analyses indicated that duplication of these segments occurred after the monocot-dicot divergence and probably after the eurosid I-eurosid II divergence. Known transposable elements were found to occur in putatively duplicated segments to a far greater extent than expected on the basis of their genome-wide distribution, suggesting that transposition may have played a role in segmental duplication in this species.     

Non-random association of transposable elements with duplicated genomic blocks in Arabidopsis thaliana

The genome of Arabidopsis thaliana is known to contain numerous open reading frames apparently encoding transposases. In order to test the hypothesis that transposable elements have played a role in segmental duplication in this species, we compared the distribution of transposable elements with that of genomic windows that shared gene families to a greater extent than expected by chance. Phylogenetic analyses indicated that duplication of these segments occurred after the monocot-dicot divergence and probably after the eurosid I–eurosid II divergence. Known transposable elements were found to occur in putatively duplicated segments to a far greater extent than expected on the basis of their genome-wide distribution, suggesting that transposition may have played a role in segmental duplication in this species.     

Role of Aromatic Amino-acid Residues in the Binding of Enzymes and Proteins to Nucleic Acids

MANY studies have been made of the specificity of interaction between nucleic acids and polypeptides, proteins and enzymes^1,2. Electrostatic forces between basic amino-acids and phosphate groups contribute to the stability of the complexes, but selective recognition requires more specific interactions which are not yet understood. The recognition of a specific region of a nucleic acid could be explained if this region has some particular conformation or if there are specific interactions between a few amino-acid residues and the bases of this region. We wish to report results which show that the aromatic amino-acids tryptophan and tyrosine can interact with nucleic acid bases in double stranded nucleic acids. They suggest that aromatic amino-acid residues of enzymes and proteins could participate in the binding to nucleic acids by intercalating between the bases and thus constraining the nucleic acid molecule to adopt a definite position with respect to the protein molecule.     

Stringency of the 2-His–1-Asp Active-Site Motif in Prolyl 4-Hydroxylase

The non-heme iron(II) dioxygenase family of enzymes contain a common 2-His–1-carboxylate iron-binding motif. These enzymes catalyze a wide variety of oxidative reactions, such as the hydroxylation of aliphatic C–H bonds. Prolyl 4-hydroxylase (P4H) is an α-ketoglutarate-dependent iron(II) dioxygenase that catalyzes the post-translational hydroxylation of proline residues in protocollagen strands, stabilizing the ensuing triple helix. Human P4H residues His412, Asp414, and His483 have been identified as an iron-coordinating 2-His–1-carboxylate motif. Enzymes that catalyze oxidative halogenation do so by a mechanism similar to that of P4H. These halogenases retain the active-site histidine residues, but the carboxylate ligand is replaced with a halide ion. We replaced Asp414 of P4H with alanine (to mimic the active site of a halogenase) and with glycine. These substitutions do not, however, convert P4H into a halogenase. Moreover, the hydroxylase activity of D414A P4H cannot be rescued with small molecules. In addition, rearranging the two His and one Asp residues in the active site eliminates hydroxylase activity. Our results demonstrate a high stringency for the iron-binding residues in the P4H active site. We conclude that P4H, which catalyzes an especially demanding chemical transformation, is recalcitrant to change.     

Presence of a characteristic D-D-E motif in IS1 transposase

Transposases encoded by various transposable DNA elements and retroviral integrases belong to a family of proteins with three conserved acidic amino acids, D, D, and E, constituting the D-D-E motif that represents the active center of the proteins. IS1, one of the smallest transposable elements in bacteria, encodes a transposase which has been thought not to belong to the family of proteins with the D-D-E motif. In this study, we found several IS1 family elements that were widely distributed not only in eubacteria but also in archaebacteria. The alignment of the transposase amino acid sequences from these IS1 family elements showed that out of 14 acidic amino acids present in IS1 transposase, three (D, D, and E) were conserved in corresponding positions in the transposases encoded by all the elements. Comparison of the IS1 transposase with other proteins with the D-D-E motif revealed that the polypeptide segments surrounding each of the three acidic amino acids were similar. Furthermore, the deduced secondary structures of the transposases encoded by IS1 family elements were similar to one another and to those of proteins with the D-D-E motif. These results strongly suggest that IS1 transposase has the D-D-E motif and thus belongs to the family of proteins with the D-D-E motif. In fact, mutant IS1 transposases with an amino acid substitution for each of the three acidic amino acids possibly constituting the D-D-E motif were not able to promote transposition of IS1, supporting this hypothesis. The D-D-E motif identified in IS1 transposase differs from those in the other proteins in that the polypeptide segment between the second D and third E in IS1 transposase is the shortest, 24 amino acids in length. Because of this difference, the presence of the D-D-E motif in IS1 transposase has not been discovered for some time.     

Functions of tetracycline efflux proteins that do not involve tetracycline

Tet(L) and Tet(K) are specific antibiotic-resistance determinants. They catalyze efflux of a tetracycline(Tc)-divalent metal complex in exchange for protons, as do other Tet efflux proteins. These Tet proteins also catalyze Na+ and K+ exchange for protons. Each of the "cytoplasmic substrates", Na+, K+ and the Tc-metal ion complex, can also be exchanged for K+, a catalytic mode that accounts for the long-recognized K+ uptake capacity conferred by some Tet proteins. The multiple catalytic modes of Tet(L) and Tet(K) provide potential new avenues for development of inhibitors of these efflux systems as well as avenues for exploration of structure-function relationships. The multiple catalytic modes of Tet(L), which is chromosomally encoded in Bacillus subtilis, also correspond to diverse physiological roles, including roles in antibiotic-, Na+-, and alkali-resistance as well as K+ acquisition. The use of K+ as an external coupling ion may contribute not only to the organism's K+ uptake capacity but also to its ability to exclude Na+ and Tc at elevated pH values. Regulation of the chromosomal tetL gene by Tc has been proposed to involve a translational re-initiation mechanism that is novel for an antibiotic-resistance gene and increases Tet expression seven-fold. Other elements of tetL expression and its regulation are already evident, including gene amplification and use of multiple promoters. However, further studies are required to clarify the full panoply of regulatory mechanisms, and their integration to ensure different levels of tetL expression that are optimal for its different functions. It will also be of interest to investigate the implications of Tet(L) and Tet(K) multifunctionality on the emergence and persistence of these antibiotic-resistance genes.