Editing of hnRNP K protein mRNA in colorectal adenocarcinoma and surrounding mucosa.

The heterogeneous nuclear ribonucleoprotein K (hnRNP K) protein is an RNA-binding protein involved in many processes that compose gene expression. K protein is upregulated in the malignant processes and has been shown to modulate the expression of genes involved in mitogenic responses and tumorigenesis. To explore the possibility that there are alternative isoforms of K protein expressed in colon cancer, we amplified and sequenced K protein mRNA that was isolated from colorectal cancers as well as from normal tissues surrounding the tumours. Sequencing revealed a single G-to-A base substitution at position 274 that was found in tumours and surrounding mucosa, but not in individuals that had no colorectal tumour. This substitution most likely reflects an RNA editing event because it was not found in the corresponding genomic DNAs. Sequencing of RNA from normal colonic mucosa of patients with prior resection of colorectal cancer revealed only the wild-type K protein transcript, indicating that G274A isoform is tumour related. To our knowledge, this is the first example of an RNA editing event in cancer and its surrounding tissue, a finding that may offer a new diagnostic and treatment marker.

Homology modeling of the CG-specific DNA methyltransferase SssI and its complexes with DNA and AdoHcy.

Prokaryotic DNA methyltransferase M.SssI recognizes and methylates C5 position of the cytosine residue within the CG dinucleotides in DNA. It is an excellent model for studying the mechanism of interaction between CG-specific eukaryotic methyltransferases and DNA. We have built a structural model of M.SssI in complex with the substrate DNA and its analogues as well as the cofactor analogue S-adenosyl-L-homocysteine (AdoHcy) using the previously solved structures of M.HhaI and M.HaeIII as templates. The model was constructed according to the recently developed "FRankenstein's monster" approach. Based on the model, amino acid residues taking part in cofactor binding, target recognition and catalysis were predicted. We also modeled covalent modification of the DNA substrate and studied its influence on protein-DNA interactions.

Mutational analysis of basic residues in the N-terminus of the rRNA:m6A methyltransferase ErmC'.

Erm methyltransferases mediate the resistance to the macrolide-lincosamide-streptogramin B antibiotics via dimethylation of a specific adenine residue in 23S rRNA. The role of positively charged N-terminal residues of the ErmC' methyltransferase in RNA binding and/or catalysis was determined. Mutational analysis of amino acids K4 and K7 was performed and the mutants were characterized in in vivo and in vitro experiments. The K4 and K7 residues were suggested not to be essential for the enzyme activity but to provide a considerable support for the catalytic step of the reaction, probably by maintaining the optimum conformation of the transition state through interactions with the phosphate backbone of RNA.

Geometric analysis of cross-linkability for protein fold discrimination.

Protein structure provides insight into the evolutionary origins, functions, and mechanisms of proteins. We are pursuing a minimalist approach to protein fold identification that characterizes possible folds in terms of consistency of their geometric features with restraints derived from relatively cheap, high-throughput experiments. One such experiment is residue-specific cross-linking analyzed by mass spectrometry. This paper presents a suite of novel lower- and upper-bounding algorithms for analyzing the distance between surface cross-link sites and thereby validating predicted models against experimental cross-linking results. Through analysis and computational experiments, using simulated and published experimental data, we demonstrate that our algorithms enable effective model discrimination.

Three-dimensional modeling of the I-TevI homing endonuclease catalytic domain, a GIY-YIG superfamily member, using NMR restraints and Monte Carlo dynamics.

Using a recent version of the SICHO algorithm for in silico protein folding, we made a blind prediction of the tertiary structure of the N-terminal, independently folded, catalytic domain (CD) of the I-TevI homing endonuclease, a representative of the GIY-YIG superfamily of homing endonucleases. The secondary structure of the I-TevI CD has been determined using NMR spectroscopy, but computational sequence analysis failed to detect any protein of known tertiary structure related to the GIY-YIG nucleases (Kowalski et al., Nucleic Acids Res., 1999, 27, 2115-2125). To provide further insight into the structure-function relationships of all GIY-YIG superfamily members, including the recently described subfamily of type II restriction enzymes (Bujnicki et al., Trends Biochem. Sci., 2000, 26, 9-11), we incorporated the experimentally determined and predicted secondary and tertiary restraints in a reduced (side chain only) protein model, which was minimized by Monte Carlo dynamics and simulated annealing. The subsequently elaborated full atomic model of the I-TevI CD allows the available experimental data to be put into a structural context and suggests that the GIY-YIG domain may dimerize in order to bring together the conserved residues of the active site.

Cloning and characterization of M.LmoA118I, a novel DNA:m4C methyltransferase from the Listeria monocytogenes phage A118, a close homolog of M.NgoMXV.

A homolog of M.NgoMXV DNA:m4C methyltransferase has been identified among the open reading frames deduced from the genomic sequence of Listeria monocytogenes phage A118 [Loessner et al., 2000]. The gene coding for this putative protein has been cloned in Escherichia coli and its enzymatic activity in vivo in this host have been analyzed. Remarkably, despite M.NgoMXV and M.LmoA118I exhibit high sequence similarity (58% identical and 19% conservatively substituted residues), their target preferences differ: both proteins exhibit "relaxed" sequence specificity, but while M.LmoA118I more efficiently methylates GGCC sites, it seems to target only a subset of CCWGG sites methylated by M.NgoMXV.

Cloning of enterohemorrhagic Escherichia coli phage VT-2 dam methyltransferase.

Enterobacterial GATC-specific DNA adenine methyltransferase (Dam) plays an essential role in regulation of DNA replication, methyl-directed mismatch repair, transposition and gene expression. In Salmonella typhimurium it has been shown to directly control virulence. In this paper we report cloning and expression of the dam gene from the Shiga toxin-producing VT2-Sa prophage of enterohemorrhagic Escherichia coli O157. Comparisons of the predicted amino acid sequence indicates that Dam methyltransferases of E. coli phages VT2-Sa, 933W, T1 and Haemophilus influenzae phage HP1 make up a separate subgroup of adenine-N6 methyltransferases. These proteins are similar to the gamma subfamily of amino-methyltransferases in respect to the linear order of sequence motifs and the presence of the hallmark "NPPY" tetrapeptide. However, they apparently lack an autonomous target-recognizing domain at the C-terminus of the catalytic domain and therefore we propose to dub them as a "mini-gamma" subfamily.

Site-directed mutagenesis defines the catalytic aspartate in the active site of the atypical DNA: m4C methyltransferase M.NgoMXV.

M.NgoMXV is one of the few atypical DNA:m4C methyltransferases that does not possess a serine residue in its predicted active site. We previously reported a homology model of M.NgoMXV and argued that the aspartate side chain at a corresponding position, similarly to some DNA:m6 A-specific enzymes, is essential for the methyltransferase activity (Radlinska et al., 1999). Here we report the corrected amino acid sequence of M.NgoMXV and the analysis of substitution of D68 with alanine or serine, which both render the enzyme totally inactive.

Pcons: a neural-network-based consensus predictor that improves fold recognition.

During recent years many protein fold recognition methods have been developed, based on different algorithms and using various kinds of information. To examine the performance of these methods several evaluation experiments have been conducted. These include blind tests in CASP/CAFASP, large scale benchmarks, and long-term, continuous assessment with newly solved protein structures. These studies confirm the expectation that for different targets different methods produce the best predictions, and the final prediction accuracy could be improved if the available methods were combined in a perfect manner. In this article a neural-network-based consensus predictor, Pcons, is presented that attempts this task. Pcons attempts to select the best model out of those produced by six prediction servers, each using different methods. Pcons translates the confidence scores reported by each server into uniformly scaled values corresponding to the expected accuracy of each model. The translated scores as well as the similarity between models produced by different servers is used in the final selection. According to the analysis based on two unrelated sets of newly solved proteins, Pcons outperforms any single server by generating approximately 8%-10% more correct predictions. Furthermore, the specificity of Pcons is significantly higher than for any individual server. From analyzing different input data to Pcons it can be shown that the improvement is mainly attributable to measurement of the similarity between the different models. Pcons is freely accessible for the academic community through the protein structure-prediction metaserver at http://bioinfo.pl/meta/.

In silico analysis of the tRNA:m1A58 methyltransferase family: homology-based fold prediction and identification of new members from Eubacteria and Archaea.

The amino acid sequences of Gcd10p and Gcd14p, the two subunits of the tRNA:(1-methyladenosine-58; m(1)A58) methyltransferase (MTase) of Saccharomyces cerevisiae, have been analyzed using iterative sequence database searches and fold recognition programs. The results suggest that the 'catalytic' Gcd14p and 'substrate binding' Gcd10p are related to each other and to a group of prokaryotic open reading frames, which were previously annotated as hypothetical protein isoaspartate MTases in sequence databases. It is predicted that the prokaryotic proteins are genuine tRNA:m(1)A MTases based on similarity of their predicted active site to the Gcd14p family. In addition to the MTase domain, an additional domain was identified in the N-terminus of all these proteins that may be involved in interaction with tRNA. These results suggest that the eukaryotic tRNA:m(1)A58 MTase is a product of gene duplication and divergent evolution of a possibly homodimeric prokaryotic enzyme.

Reassignment of specificities of two cap methyltransferase domains in the reovirus lambda 2 protein.

BACKGROUND: The reovirus lambda2 protein catalyzes mRNA capping, that is, addition of a guanosine to the 5' end of each transcript in a 5'-to-5' orientation, as well as transfer of a methyl group from S-adenosyl-L-methionine (AdoMet) to the N7 atom of the added guanosyl moiety and subsequently to the ribose 2'-O atom of the first template-encoded nucleotide. The structure of the human reovirus core has been solved at 3.6 A resolution, revealing a series of domains that include a putative guanylyltransferase domain and two putative methyltransferase (MTase) domains. It has been suggested that the order of domains in the lambda2 protein corresponds to the order of reactions in the pathway and that the m7G (cap 0) and the 2'-O-ribose (cap 1) MTase activities may be exerted by the MTase 1 and the MTase 2 domains, respectively. RESULTS: We show that the reovirus MTase 1 domain shares a putative active site with the structurally characterized 2'-O-ribose MTases, including vaccinia virus cap 1 MTase, whereas the MTase 2 domain is structurally similar to glycine N-MTase. CONCLUSIONS: On the basis of our analysis of the structural details we propose that the previously suggested functional assignments of the MTase 1 and MTase 2 domains should be swapped.

Sequence analysis and structure prediction of aminoglycoside-resistance 16S rRNA:m7G methyltransferases.

Methylation of G1405 within bacterial 16S ribosomal RNA results in high-level resistance to specific combinations of aminoglycoside antibiotics. Only a few closely related methyltransferases (MTases), which carry out the respective modification (here dubbed "Agr", for aminoglycoside resistance), are known. It is not clear, whether they are related to "typical" S-adenosylmethionine (AdoMet)-dependent MTases or not. Demydchuk et al., 1998 proposed that the cofactor-binding region is localized at the C-terminus of Agr MTases, which implies an interesting case of sequence permutation. Since the Agr MTases lack significant sequence similarity to other proteins, we tested that hypothesis using more sensitive sequence/structure threading approach. Structure prediction confirmed the presence of a putative AdoMet-binding site in these proteins, albeit at a distinct location, resembling that of "typical", non-permuted MTases. Additionally, a small alpha-helical domain dissimilar to other proteins in the database was identified in the N-terminal region of Agr MTases. Comparison of a three-dimensional model of the Agr family member with a recently solved structure of reovirus mRNA capping MTase suggests that the mechanism of guanine-N7 methylation in rRNA and mRNA may be different.

Structure prediction meta server.

UNLABELLED: The Structure Prediction Meta Server offers a convenient way for biologists to utilize various high quality structure prediction servers available worldwide. The meta server translates the results obtained from remote services into uniform format, which are consequently used to request a jury prediction from a remote consensus server Pcons. AVAILABILITY: The structure prediction meta server is freely available at http://BioInfo.PL/meta/, some remote servers have however restrictions for non-academic users, which are respected by the meta server. SUPPLEMENTARY INFORMATION: Results of several sessions of the CAFASP and LiveBench programs for assessment of performance of fold-recognition servers carried out via the meta server are available at http://BioInfo.PL/services.html.

mRNA:guanine-N7 cap methyltransferases: identification of novel members of the family, evolutionary analysis, homology modeling, and analysis of sequence-structure-function relationships.

BACKGROUND: The 5'-terminal cap structure plays an important role in many aspects of mRNA metabolism. Capping enzymes encoded by viruses and pathogenic fungi are attractive targets for specific inhibitors. There is a large body of experimental data on viral and cellular methyltransferases (MTases) that carry out guanine-N7 (cap 0) methylation, including results of extensive mutagenesis. However, a crystal structure is not available and cap 0 MTases are too diverged from other MTases of known structure to allow straightforward homology-based interpretation of these data. RESULTS: We report a 3D model of cap 0 MTase, developed using sequence-to-structure threading and comparative modeling based on coordinates of the glycine N-methyltransferase. Analysis of the predicted structural features in the phylogenetic context of the cap 0 MTase family allows us to rationalize most of the experimental data available and to propose potential binding sites. We identified a case of correlated mutations in the cofactor-binding site of viral MTases that may be important for the rational drug design. Furthermore, database searches and phylogenetic analysis revealed a novel subfamily of hypothetical MTases from plants, distinct from "orthodox" cap 0 MTases. CONCLUSIONS: Computational methods were used to infer the evolutionary relationships and predict the structure of Eukaryotic cap MTase. Identification of novel cap MTase homologs suggests candidates for cloning and biochemical characterization, while the structural model will be useful in designing new experiments to better understand the molecular function of cap MTases.

Unusual evolutionary history of the tRNA splicing endonuclease EndA: relationship to the LAGLIDADG and PD-(D/E)XK deoxyribonucleases.

The tRNA splicing endoribonuclease EndA from Methanococcus jannaschii is a homotetramer formed via heterologous interaction between the two pairs of homodimers. Each monomer consists of two alpha/beta domains, the N-terminal domain (NTD) and the C-terminal domain (CTD) containing the RNase A-like active site. Comparison of the EndA coordinates with the publicly available protein structure database revealed the similarity of both domains to site-specific deoxyribonucleases: the NTD to the LAGLIDADG family and the CTD to the PD-(D/E)XK family. Superposition of the NTD on the catalytic domain of LAGLIDADG homing endonucleases allowed a suggestion to be made about which amino acid residues of the tRNA splicing nuclease might participate in formation of a presumptive cryptic deoxyribonuclease active site. On the other hand, the CTD and PD-(D/E)XK endonucleases, represented by restriction enzymes and a phage lambda exonuclease, were shown to share extensive similarities of the structural framework, to which entirely different active sites might be attached in two alternative locations. These findings suggest that EndA evolved from a fusion protein with at least two distinct endonuclease activities: the ribonuclease, which made it an essential "antitoxin" for the cells whose RNA genes were interrupted by introns, and the deoxyribonuclease, which provided the means for homing-like mobility. The residues of the noncatalytic CTDs from the positions corresponding to the catalytic side chains in PD-(D/E)XK deoxyribonucleases map to the surface at the opposite side to the tRNA binding site, for which no function has been implicated. Many restriction enzymes from the PD-(D/E)XK superfamily might have the potential to maintain an additional active or binding site at the face opposite the deoxyribonuclease active site, a property that can be utilized in protein engineering.

Identification of a PD-(D/E)XK-like domain with a novel configuration of the endonuclease active site in the methyl-directed restriction enzyme Mrr and its homologs.

The Escherichia coli K-12 restriction enzyme Mrr recognizes and cleaves N6-methyladenine- and 5-methylcytosine-containing DNA. Its amino acid sequence has been subjected to structure prediction and comparison with other sequences from publicly available sources. The results obtained suggest that Mrr and related putative endonucleases possess a cleavage domain typical for all the so far structurally characterized type II restriction enzymes, however with an unusual glutamine residue at the central position of the (D/E)-(D/E)XK hallmark of the active site. The "missing" acidic side chain was instead found anchored in a different, unusual position, suggesting that Mrr represents a third topological variant of the endonuclease active site in addition to the two alternatives determined previously (Skirgaila et al., 1998. J. Mol. Biol. 279, 473-481). One of the newly identified putative endonucleases from the Mrr family is composed of two diverged cleavage domains, which possess both the "typical" D-EXK and the "Mrr-like" D-QXK variants of the active site. Among the Mrr homologs there are also proteins from yeast, in which restriction phenotype has not been observed, suggesting that the free-standing Eukaryotic PD-(D/E)XK superfamily members might be implicated in other cellular processes involving enzymatic DNA cleavage.

The herpesvirus alkaline exonuclease belongs to the restriction endonuclease PD-(D/E)XK superfamily: insight from molecular modeling and phylogenetic analysis.

The PD-(D/E)XK superfamily of deoxyribonucleases (ENases) comprises restriction endonucleases, exonucleases and nicking enzymes, which share a common fold and the architecture of the active site. Their extreme divergence generally hampers identification of novel members based solely on sequence comparisons. Here we report a remote similarity between the phage lambda exonuclease (lambda-exo), branching out early in the evolutionary history of ENases (3), with the family of alkaline exonucleases (AE) encoded by various viruses infecting higher Eukaryota. The predicted structural compatibility and the conservation of the functionally important residues between AE and ENases strongly suggest a distant evolutionary relationship between these proteins. According to the results of extensive sequence database mining, sequence/structure threading and molecular modeling it is plausible that the AE proteins with lambda-exo and some other putative phage-encoded exonucleases form a distinct subfamily of PD-(D/E)XK ENases. The phylogenetic history of this subfamily is inferred using sequence alignment and distance matrix methods.

LiveBench-1: continuous benchmarking of protein structure prediction servers.

We present a novel, continuous approach aimed at the large-scale assessment of the performance of available fold-recognition servers. Six popular servers were investigated: PDB-Blast, FFAS, T98-lib, GenTHREADER, 3D-PSSM, and INBGU. The assessment was conducted using as prediction targets a large number of selected protein structures released from October 1999 to April 2000. A target was selected if its sequence showed no significant similarity to any of the proteins previously available in the structural database. Overall, the servers were able to produce structurally similar models for one-half of the targets, but significantly accurate sequence-structure alignments were produced for only one-third of the targets. We further classified the targets into two sets: easy and hard. We found that all servers were able to find the correct answer for the vast majority of the easy targets if a structurally similar fold was present in the server's fold libraries. However, among the hard targets--where standard methods such as PSI-BLAST fail--the most sensitive fold-recognition servers were able to produce similar models for only 40% of the cases, half of which had a significantly accurate sequence-structure alignment. Among the hard targets, the presence of updated libraries appeared to be less critical for the ranking. An "ideally combined consensus" prediction, where the results of all servers are considered, would increase the percentage of correct assignments by 50%. Each server had a number of cases with a correct assignment, where the assignments of all the other servers were wrong. This emphasizes the benefits of considering more than one server in difficult prediction tasks. The LiveBench program (http://BioInfo.PL/LiveBench) is being continued, and all interested developers are cordially invited to join.

Grouping together highly diverged PD-(D/E)XK nucleases and identification of novel superfamily members using structure-guided alignment of sequence profiles.

The PD-(D/E)XK nuclease domains, initially identified in type II restriction enzymes, serve as models for studying aspects of protein-DNA interactions, mechanisms of phosphodiester hydrolysis, and provide indispensable tools for techniques in genetic engineering and molecular medicine. However, the low degree of amino acid conservation hampers the possibility of identification of PD-(D/E)XK superfamily members based solely on sequence comparisons. In several proteins implicated in DNA recombination and repair the restriction enzyme-like nuclease domain has been found only after the corresponding structures were determined experimentally. Here, we identified highly diverged variants of the PD-(D/E)XK domain in many proteins and open reading frames using iterative database searches and progressive, structure-guided alignment of sequence profiles. We predicted the possible cellular function for many hypothetical proteins based on their relative similarity to characterized nucleases or observed presence of additional domains. We also identified the nuclease domain in genuine recombinases and restriction enzymes, whose homology to other PD-(D/E)XK enzymes has not been demonstrated previously. The first superfamily-wide comparative analysis, not limited to nucleases of known structure, will guide cloning and characterization of novel enzymes and planning new experiments to better understand those already studied.

Polyphyletic evolution of type II restriction enzymes revisited: two independent sources of second-hand folds revealed.

Using algorithms for protein sequence analysis we predict that some of the canonical type II and type IIS restriction enzymes have an active site with a substantially different architecture and fold from the "typical" PD-(D/E)xK superfamily. These results suggest that they are related to nucleases from the HNH and GIY-YIG superfamilies.