DNA methyltransferases.

Mammals have long been known to tag their DNA by the addition of methyl groups to cytosine residues. Only quite recently, however, has the functional significance of DNA methylation established a firm footing. Evidence now indicates that DNA methylation is essential for development, and is involved in both programmed and ectopic gene inactivation. Recent structural and mechanistic work on bacterial cytosine-5-methyltransferases has provided much insight into the function of the carboxy-terminal catalytic domain of eukaryotic cytosine-5-methyltransferases; evidence is emerging that the amino-terminal domain targets the enzyme to the replication machinery and may be involved in sensing the pre-existing methylation state of the DNA.

Structure of a covalently trapped catalytic complex of HIV-1 reverse transcriptase: implications for drug resistance.

A combinatorial disulfide cross-linking strategy was used to prepare a stalled complex of human immunodeficiency virus-type 1 (HIV-1) reverse transcriptase with a DNA template:primer and a deoxynucleoside triphosphate (dNTP), and the crystal structure of the complex was determined at a resolution of 3.2 angstroms. The presence of a dideoxynucleotide at the 3'-primer terminus allows capture of a state in which the substrates are poised for attack on the dNTP. Conformational changes that accompany formation of the catalytic complex produce distinct clusters of the residues that are altered in viruses resistant to nucleoside analog drugs. The positioning of these residues in the neighborhood of the dNTP helps to resolve some long-standing puzzles about the molecular basis of resistance. The resistance mutations are likely to influence binding or reactivity of the inhibitors, relative to normal dNTPs, and the clustering of the mutations correlates with the chemical structure of the drug.

Induced alpha helix in the VP16 activation domain upon binding to a human TAF.

Activation domains are functional modules that enable sequence-specific DNA binding proteins to stimulate transcription. The structural basis for the function of activation domains is poorly understood. A combination of nuclear magnetic resonance (NMR) and biochemical experiments revealed that the minimal acidic activation domain of the herpes simplex virus VP16 protein undergoes an induced transition from random coil to alpha helix upon binding to its target protein, hTAFII31 (a human TFIID TATA box-binding protein-associated factor). Identification of the two hydrophobic residues that make nonpolar contacts suggests a general recognition motif of acidic activation domains for hTAFII31.

Repair of DNA methylphosphotriesters through a metalloactivated cysteine nucleophile.

The Escherichia coli Ada protein repairs methylphosphotriesters in DNA by direct, irreversible methyl transfer to one of its own cysteines. Upon methyl transfer, Ada acquires the ability to bind specific DNA sequences and thereby to induce genes that confer resistance to methylating agents. The amino-terminal domain of Ada, which comprises the methylphosphotriester repair and sequence-specific DNA binding elements, contains a tightly bound zinc ion. Analysis of the zinc binding site by cadmium-113 nuclear magnetic resonance and site-directed mutagenesis revealed that zinc participates in the autocatalytic activation of the active site cysteine and may also function as a conformational switch.

Isolation and structure of a covalent cross-link adduct between mitomycin C and DNA.

A DNA cross-link adduct of the antitumor agent mitomycin C (MC) to DNA has been isolated and characterized; the results provide direct proof for bifunctional alkylation of DNA by MC. Exposure of MC to Micrococcus luteus DNA under reductive conditions and subsequent nuclease digestion yielded adducts formed between MC and deoxyguanosine residues. In addition to the two known monoadducts, a bisadduct was obtained. Reductive MC activation with Na2S2O4 (sodium dithionite) leads to exclusive bifunctional alkylation. The structure of the bisadduct was determined by spectroscopic methods that included proton magnetic resonance, differential Fourier transform infrared spectroscopy, and circular dichroism. Formation of the same bisadduct in vivo was demonstrated upon injection of rats with MC. Computer-generated models of the bisadduct that was incorporated into the center of the duplex B-DNA decamer d(CGTACGTACG)2 indicated that the bisadduct fit snugly into the minor groove with minimal distortion of DNA structure. A mechanistic analysis of the factors that govern monofunctional and bifunctional adduct formation is presented.

A mammalian DNA repair enzyme that excises oxidatively damaged guanines maps to a locus frequently lost in lung cancer.

BACKGROUND: Guanine residues in the genome are vulnerable to attack by free radicals and reactive oxygen species. A major lesion thus produced, 8-oxoguanine (OG), causes mutations by mis-pairing with adenine during replication. In bacteria and budding yeast, OG is removed from the genome through the action of base-excision DNA repair (BER) enzymes, which catalyze expulsion of the aberrant base and excision of its sugar moiety from the DNA backbone. Although OG is known to be produced in and cleansed from mammalian genomes, the enzymes responsible for OG repair in these cells have remained elusive. RESULTS: Here, we report the cloning and biochemical characterization of mammalian BER enzymes that specifically target OG residues in DNA. These 8-oxoguanine DNA glycosylases, hOgg1 (human) and mOgg1 (murine), are homologous to each other and to yeast Ogg1. They also contain an active site motif - the Helix-hairpin-Helix, Gly/Pro-rich-Asp motif - characteristic of a superfamily of BER proteins with a similar core fold and active site geometry. Both hOgg1 and mOgg1 exhibit exquisite selectivity for the base opposite OG in DNA, operating with high efficiency only on OG base-paired to cytosine. Furthermore, hOgg1 and mOgg1 are unable to process a panel of alternative lesions, including 8-oxoadenine, yet bind with high affinity to synthetic abasic site analogs. The proteins operate through a classical glycosylase/lyase catalytic mechanism; mutation of a catalytically essential lysine residue results in loss of catalytic potency but retention of binding to OG-containing oligonucleotides. The hOGG1 gene is localized on the short arm of chromosome 3 (3p25/26) in a region commonly deleted in cancers. CONCLUSIONS: These results conclusively establish the existence and identity of an 8-oxoguanine DNA glycosylase/lyase in human and murine cells, completing the triad of proteins that together protect mammals from the genotoxic effects of guanine oxidation. The observation that at least one allele of hOGG1 is commonly deleted in cancer cells suggests that such cells may possess a reduced capacity to counter the mutagenic effects of reactive oxygen species, a deficiency that could increase their overall genomic instability. This speculation is fueled by recent observations that cells constitutively active for the Ras/Raf pathway constitutively produce high levels of superoxide, a known generator of OG.

Repair of oxidatively damaged guanine in Saccharomyces cerevisiae by an alternative pathway.

BACKGROUND: Transversion mutations are caused by 8-oxoguanine (OG), a DNA lesion produced by the spontaneous oxidation of guanine nucleotides, which mis-pairs with adenine during replication. Resistance to this mutagenic threat is mediated by the GO system, the components of which are functionally conserved in bacteria and mammals. To date, only one of three GO system components has been identified in the budding yeast Saccharomyces cerevisiae, namely the OG:C-specific glycosylase/lyase yOgg1. Furthermore, S. cerevisiae has been reported to contain a unique glycosylase/lyase activity, yOgg2, which excises OG residues opposite adenines. Paradoxically, according to the currently accepted model, yOgg2 activity should increase the mutagenicity of OG lesions. Here we report the isolation of yOgg2 and the elucidation of its role in oxidative mutagenesis. RESULTS: Borohydride-dependent cross-linking using an OG-containing oligonucleotide substrate led to the isolation of yOgg1 and a second protein, Ntg1, which had previously been shown to process oxidized pyrimidines in DNA. We demonstrate that Ntg1 has OG-specific glycosylase/lyase activity indistinguishable from that of yOgg2. Targeted disruption of the NTG1 gene resulted in complete loss of yOgg2 activity and yeast lacking NTG1 had an elevated rate of A:T to C:G transversions. CONCLUSIONS: The Ntg1 and yOgg2 activities are encoded by a single gene. We propose that yOgg2 has evolved to process OG:A mis-pairs that have arisen through mis-incorporation of 8-oxo-dGTP during replication. Thus, the GO system in S. cerevisiae is fundamentally distinct from that in bacteria and mammals.

Identification of a new uracil-DNA glycosylase family by expression cloning using synthetic inhibitors.

BACKGROUND: The cellular environment exposes DNA to a wide variety of endogenous and exogenous reactive species that can damage DNA, thereby leading to genetic mutations. DNA glycosylases protect the integrity of the genome by catalyzing the first step in the base excision-repair of lesions in DNA. RESULTS: Here, we report a strategy to conduct genome-wide screening for expressed DNA glycosylases, based on their ability to bind to a library of four synthetic inhibitors that target the enzyme's active site. These inhibitors, used in conjunction with the in vitro expression cloning procedure, led to the identification of novel Xenopus and human proteins, xSMUG1 and hSMUG1, respectively, that efficiently excise uracil residues from DNA. Despite a lack of statistically significant overall sequence similarity to the two established classes of uracil-DNA glycosylases, the SMUG1 enzymes contain motifs that are hallmarks of a shared active-site structure and overall protein architecture. The unusual preference of SMUG1 for single-stranded rather than double-stranded DNA suggests a unique biological function in ridding the genome of uracil residues, which are potent endogenous mutagens. CONCLUSIONS: The 'proteomics' approach described here has led to the isolation of a new family of uracil-DNA glycosylases. The three classes of uracil-excising enzymes (SMUG1 being the most recently discovered) represent a striking example of structural and functional conservation in the almost complete absence of sequence conservation.

Only one of the two DNA-bound orientations of AP-1 found in solution cooperates with NFATp.

BACKGROUND: The transcription factor AP-1 activates the expression of numerous genes in response to mitogenic stimuli. AP-1 regulates gene expression both through solitary binding to independent recognition sites and, in cooperation with various heterologous transcription factors, through targeting to composite response elements. The two subunits that make up the AP-1 heterodimer, Fos and Jun, possess identical residues at positions that make sequence-specific contacts to DNA. This degeneracy leaves the protein with no apparent way of orienting itself uniquely on DNA by differentially recognizing its two non-identical half-sites. Here, we have analyzed the orientation of the AP-1 basic-leucine-zipper (bZip) domain on a cognate site, both alone and in the cooperative complex formed together with the 'nuclear factor of activated T cells' (NFATp). RESULTS: The results of affinity cleaving experiments demonstrate that, in solution, the AP-1 bZip binds DNA as a mixture of two orientational isomers. However, in the cooperative complex formed with NFATp on a composite response element, the AP-1 bZip adopts a single orientation, with Jun and Fos bound to the NFATp-proximal and NFATp-distal half-sites, respectively. Protein cross-linking experiments demonstrate that protein-protein contacts are responsible for this 'orientational locking'. CONCLUSIONS: Our results demonstrate that, through protein-protein interactions, one protein can force another to adopt a single DNA-bound orientation. Thus, cooperative interactions between adjacent regulatory proteins can influence not only the energetics of their interactions with DNA, but also their precise geometric and stereochemical arrangement. Because orientational isomers present markedly different structures to the transcriptional apparatus, it seems likely that orientation will exert an effect on the ability to activate transcription.

Cloning of a yeast 8-oxoguanine DNA glycosylase reveals the existence of a base-excision DNA-repair protein superfamily.

BACKGROUND: Reactive oxygen species, ionizing radiation, and other free radical generators initiate the conversion of guanine (G) residues in DNA to 8-oxoguanine (OG), which is highly mutagenic as it preferentially mispairs with adenine (A) during replication. Bacteria counter this threat with a multicomponent system that excises the lesion, corrects OG:A mispairs and cleanses the nucleotide precursor pool of dOGTP. Although biochemical evidence has suggested the existence of base-excision DNA repair proteins specific for OG in eukaryotes, little is known about these proteins. RESULTS: Using substrate-mimetic affinity chromatography followed by a mechanism-based covalent trapping procedure, we have isolated a base-excision DNA repair protein from Saccharomyces cerevisiae that processes OG opposite cytosine (OG:C) but acts only weakly on OG:A. A search of the yeast genome database using peptide sequences from the protein identified a gene, OGG1, encoding a predicted 43 kDa (376 amino acid) protein, identical to one identified independently by complementation cloning. Ogg1 has OG:C-specific base-excision DNA repair activity and also intrinsic beta-lyase activity, which proceeds through a Schiff base intermediate. Targeted disruption of the OGG1 gene in yeast revealed a second OG glycosylase/lyase protein, tentatively named Ogg2, which differs from Ogg1 in that it preferentially acts on OG:G. CONCLUSIONS: S. cerevisiae has two OG-specific glycosylase/lyases, which differ significantly in their preference for the base opposite the lesion. We suggest that one of these, Ogg1, is closely related in overall three-dimensional structure to Escherichia coli endonuclease III (endo III), a glycosylase/lyase that acts on fragmented and oxidatively damaged pyrimidines. We have recently shown that AlkA, a monofunctional DNA glycosylase that acts on alkylated bases, is structurally homologous to endo III. We have now identified a shared active site motif amongst these three proteins. Using this motif as a protein database searching tool, we find that it is present in a number of other base-excision DNA repair proteins that process diverse lesions. Thus, we propose the existence of a DNA glycosylase superfamily, members of which possess a common fold yet act upon remarkably diverse lesions, ranging from UV photoadducts to mismatches to alkylated or oxidized bases.

A DNA-bending protein interacts with an essential upstream regulatory element of the human embryonic beta-like globin gene.

The mammalian beta-like globin gene family has served as an important model system for analysis of tissue- and developmental state-specific gene regulation. Although the activities of a number of regulatory proteins have been implicated in the erythroid cell-specific transcription of globin genes, the mechanisms that restrict their expression to discrete stages of development are less well understood. We have previously identified a novel regulatory element (PRE II) upstream from the human embryonic beta-like globin gene (epsilon) that synergizes with other sequences to confer tissue- and stage-specific expression on a minimal epsilon-globin gene promoter in cultured embryonic erythroid cells. Binding of an erythroid nuclear protein (PRE II-binding factor [PRE-IIBF]) to the PRE II control element is required for promoter activation. Here we report on some of the biochemical properties of PREIIBF, including the characterization of its specificity and affinity for DNA. The embryonic and adult forms of PREIIBF recognize their cognate sequences with identical specificities, supporting our earlier conclusion that they are very similar proteins. PREIIBF binds DNA as a single polypeptide with an Mr of approximately 80,000 to 85,000 and introduces a bend into the target DNA molecule. These results suggest a mechanism by which PREIIBF may contribute to the regulation of the embryonic beta-like globin gene within the context of a complex locus.

The Rel family of eukaryotic transcription factors.

Members of the Rel transcription factor family mediate the response of eukaryotic cells to a broad range of environmental threats, in addition to serving an essential role in the development of certain vertebrate and insect cells. It is now apparent that there are two classes of Rel proteins, which differ in whether they bind DNA as monomers or dimers and which use markedly different mechanisms to transduce intracellular signals. Recent progress has been made towards understanding the structural basis for the fascinating biology of these proteins.

Chemical approaches toward understanding base excision DNA repair.

Despite the importance of DNA repair in protecting the genome, the molecular basis for damage recognition and repair remains poorly understood. In the base excision repair pathway (BER), DNA glycosylases recognize and excise damaged bases from DNA. This review focuses on the recent development of chemical approaches that have been applied to the study of BER enzymes. Several distinctive classes of noncleavable substrate analogs that form stable complexes with DNA glycosylases have recently been designed and synthesized. These analogs have been used for biochemical and structural analyses of protein-DNA complexes involving DNA glycosylases, and for the isolation of a novel DNA glycosylase. An approach to trap covalently a DNA glycosylase-intermediate complex has also been used to elucidate the mechanism of DNA glycosylases.

Crystallization and preliminary crystallographic analysis of a DNA (cytosine-5)-methyltransferase from Haemophilus aegyptius bound covalently to DNA.

A DNA (cytosine)-5-methyltransferase from Haemophilus aegyptius (M.Hae III), which catalyzes methyl transfer from S-adenosyl-L-methionine to DNA, has been crystallized as a covalent complex with a suicide oligonucleotide substrate. Crystals of the co-complex were grown by vapor diffusion with hanging droplets, using polyethylene glycol 3500 as the precipitant. The crystals belong to the orthorhombic space group P2(1)2(1)2(1); the unit cell parameters are a = 57.6 A, b = 108.0 A, c = 155.8 A with two protein-DNA complexes in the asymmetric unit. Complete sets of native and derivative data have been collected to 2.7 A using a laboratory source.

DNA (cytosine-5)-methyltransferases in mouse cells and tissues. Studies with a mechanism-based probe.

The mechanisms that establish and maintain methylation patterns in the mammalian genome are very poorly understood, even though perturbations of methylation patterns lead to a loss of genomic imprinting, ectopic X chromosome inactivation, and death of mammalian embryos. A family of sequence-specific DNA methyltransferases has been proposed to be responsible for the wave of de novo methylation that occurs in the early embryo, although no such enzyme has been identified. A universal mechanism-based probe for DNA (cytosine-5)-methyltransferases was used to screen tissues and cell types known to be active in de novo methylation for new species of DNA methyltransferase. All identifiable de novo methyltransferase activity was found to reside in Dnmt1. As this enzyme is the predominant de novo methyltransferase at all developmental stages inspected, it does not fit the definition of maintenance methyltransferase or hemimethylase. Recent genetic data indicate that de novo methylation of retroviral DNA in embryonic stem cells is likely to involve one or more additional DNA methyltransferases. Such enzymes were not detected and are either present in very small amounts or are very different from Dnmt1. A new method was developed and used to determine the sequence specificity of intact Dnmt1 in whole-cell lysates. Specificity was found to be confined to the sequence 5'-CpG-3'; there was little dependence on sequence context or density of CpG dinucleotides. These data suggest that any sequence-specific de novo methylation mediated by Dnmt1 is either under the control of regulatory factors that interact with Dnmt1, or is cued by alternative secondary structures in DNA.

How do DNA repair proteins locate damaged bases in the genome?

The genome is susceptible to the attack of reactive species that chemically modify the bases of DNA. If genetic information is to be transmitted faithfully to successive generations, it is essential that the genome be repaired. All organisms express proteins specifically dedicated to this task. But how do these proteins find the aberrant bases amongst the enormous number of normal ones?

The effects of N7-methylguanine on duplex DNA structure.

BACKGROUND: Non-enzymatic methylation of DNA by endogenous and exogenous agents produces a variety of adducts, of which the predominant one is N7-methyl-2'-deoxyguanosine (m7dG). Although it is known that living organisms counter the deleterious effects of m7dG by producing adduct-specific DNA repair proteins, the molecular basis for specific recognition and catalysis by these proteins is poorly understood. In addition to its role as an endogenous DNA adduct, m7dG is also widely used as an in vitro probe of protein-DNA interactions. We set out to examine whether incorporation of m7dG into DNA affects duplex DNA structure. RESULTS: We carried out a large-scale synthesis of a dodecamer containing the m7dG adduct at a single, defined position. Because the instability of m7dG precludes its incorporation into oligonucleotides by standard solid-phase methods, a novel strategy employing chemical and enzymatic synthesis was used. Characterization of the m7dG-containing dodecamer by NMR reveals no structural distortion; indeed, m7dG appears to encourage a modest shift toward a more characteristically B-form duplex. CONCLUSIONS: These results argue strongly against induced DNA distortion as a mechanism for specific recognition of m7dG by adduct-specific repair proteins. The broad substrate specificity of these repair proteins disfavors a model involving direct recognition of aberrantly placed methyl groups; hence, it may be that m7dG is recognized indirectly, perhaps by its effects on the dynamics of DNA. On the other hand, the evidence presented here suggests that m7dG interferes directly with sequence-specific recognition by DNA-binding proteins by steric blockage or by masking of required contact functionalities. The synthetic methodology used here should be generally applicable to high-resolution structural studies of oligonucleotides bearing adducts that are unstable to the conditions of solid-phase DNA synthesis.

Synthesis and biochemical evaluation of RNA containing an intrahelical disulfide crosslink.

BACKGROUND: Several factors impede the elucidation of RNA structure and function by X-ray and NMR methods, including the complexity of folded RNA motifs, the tendency of RNA to aggregate, and its ability to fold into multiple isomeric structures. The ability to constrain the process of RNA folding to give a single, homogeneous product would assist these investigations. We therefore set out to develop a synthetic procedure for the site-specific insertion of a disulfide crosslink into oligoribonucleotides. We also examined the ability of a crosslinked species to serve as a substrate for ricin, an RNA glycosylase. RESULTS: A convertible nucleoside derivative (C) suitable for the site-specific introduction of N4-alkylcytidine residues into RNA has been developed. The corresponding C phosphoramidite was employed in the synthesis of an 8-mer oligonucleotide, 5'-CGGA-GACG-3', which was then efficiently converted to an 8-mer containing two S-protected N4-(2-thioethyl)C residues. Upon deprotection and air oxidation, the 8-mer efficiently formed an intramolecular disulfide bond, yielding a GAGA tetraloop presented on a two-base-pair CpG disulfide crosslinked ministem. We show that this ministem-loop is an excellent substrate for ricin. Control 8-mers lacking the disulfide crosslink were substantially poorer substrates for ricin. CONCLUSIONS: The nucleoside chemistry described here should be generally useful for the site-specific introduction of a range of non-native functional groups into RNA. We have used this chemistry to constrain an RNA ministem through introduction of an intrahelical disulfide crosslink. That this tetraloop substrate linked to a two base-pair ministem is efficiently processed by ricin is clear evidence that ricin makes all of its energetically favorable contacts to the extreme end of the stem-loop structure, and that the two base pairs of the stem abutting the loop remain intact during recognition and processing by ricin.

The leucine zipper domain controls the orientation of AP-1 in the NFAT.AP-1.DNA complex.

BACKGROUND: Heterologous transcription factors bound to adjacent sites in eukaryotic promoters often exhibit cooperative behavior. In most instances, the molecular basis for this cooperativity is poorly understood. Our efforts have been directed toward elucidation of the mechanism of cooperativity between NFAT and AP-1, two proteins that coordinately direct expression of the T-cell growth factor interleukin-2 (IL-2). RESULTS: We have previously shown that NFAT1 orients the two subunits of AP-1, c-Jun and c-Fos, on DNA through direct protein-protein interactions. In the present study, we have constructed cJun-cFos chimeric proteins and determined their orientation using a novel affinity-cleavage technology based on chemical ligation. We find that, in the presence of NFAT, the chimeric heterodimer binds in such a way as to preserve the orientation of the AP-1 leucine zipper, but not that of the basic region. CONCLUSIONS: Protein-protein interactions between NFAT and the leucine zipper of AP-1 enable the two proteins to bind DNA cooperatively and coordinately regulate the IL-2 promoter. The chemical ligation technology presented here provides a powerful strategy for affinity cleavage studies, including those using recombinant proteins.

Trapping of a catalytic HIV reverse transcriptase*template:primer complex through a disulfide bond.

BACKGROUND: HIV-1 reverse transcriptase (RT) is a major target for the treatment of acquired immunodeficiency syndrome (AIDS). Resistance mutations in RT compromise treatment, however. Efforts to understand the enzymatic mechanism of RT and the basis for mutational resistance to anti-RT drugs have been hampered by the failure to crystallize a catalytically informative RT-substrate complex. RESULTS: We present here experiments that allow us to understand the reason for the failure to crystallize such a complex. Based on this understanding, we have devised a new approach for using a combinatorial disulfide cross-linking strategy to trap a catalytic RT*template:primer*dNTP ternary complex, thereby enabling the growth of co-crystals suitable for high-resolution structural analysis. The crystals contain a fully assembled active site poised for catalysis. The cross-link itself appears to be conformationally mobile, and the surrounding region is undistorted, suggesting that the cross-link is a structurally passive device that merely acts to prevent dissociation of the catalytic complex. CONCLUSIONS: The new strategy discussed here has resulted in the crystallization and structure determination of a catalytically relevant RT*template:primer*dNTP complex. The structure has allowed us to analyze possible causes of drug resistance at the molecular level. This information will assist efforts to develop new classes of nucleoside analog inhibitors, which might help circumvent current resistance profiles. The covalent trapping strategy described here may be useful with other protein-DNA complexes that have been refractory to structural analysis.