Neuroevolution and complexifying genetic architectures for memory and control tasks

The way genes are interpreted biases an artificial evolutionary system towards some phenotypes. When evolving artificial neural networks, methods using direct encoding have genes representing neurons and synapses, while methods employing artificial ontogeny interpret genomes as recipes for the construction of phenotypes. Here, a neuroevolution system (neuroevolution with ontogeny or NEON) is presented that can emulate a well-known neuroevolution method using direct encoding (neuroevolution of augmenting topologies or NEAT), and therefore, can solve the same kinds of tasks. Performance on challenging control and memory benchmark tasks is reported. However, the encoding used by NEON is indirect, and it is shown how characteristics of artificial ontogeny can be introduced incrementally in different phases of evolutionary search.     

Cloning and chromosomal mapping of mouse ADAM11, ADAM22 and ADAM23.

A cellular disintegrin, also called MDC and ADAM is a recently discovered gene family that encodes protein with disintegrin-like and metalloprotease-like domains. We have reported the identification of human cDNAs encoding novel ADAM family proteins that we named MDC2 and MDC3 because of their structural similarity to the MDC (Sagane, K. et al., 1998. Biochem. J. 334, 93-98). The Human Gene Nomenclature Committee assigned the gene symbols ADAM11 for the MDC, ADAM22 for the MDC2 and ADAM23 for the MDC3. Here we report the isolation of three novel murine cDNAs encoding the proteins closely related to the human ADAM11, ADAM22 and ADAM23. Their chromosomal locations in the mouse were identified by interspecies backcross mapping. The loci of these murine ADAM genes were in good accordance with the location of each human ortholog, ADAM11, ADAM22 and ADAM23. These findings suggest that three murine cDNAs that we have isolated are the murine ADAM11, ADAM22 and ADAM23 cDNAs. Northern blot analysis shows that all of these three murine ADAMs were highly expressed in the mouse brain. The structures of these ADAM proteins strongly suggest that they could function as integrin receptors. The implications of the cellular disintegrins in neural development are discussed.     

Yeast species recognition from gene sequence analyses and other molecular methods

This review discusses DNA-based methods used for identification of yeasts. Nuclear DNA reassociation was the first quantitative molecular method employed for recognition of yeast species and has provided a baseline for interpretation of other molecular comparisons. Among these, gene sequencing is the most definitive method, with ribosomal RNA gene sequences providing the preponderance of available data. Multigene analyses that include the sequences of protein encoding genes are being increasingly developed to provide a more definitive resolution of species. A number of rapid identification methods, such as denaturing gradient gel electrophoresis (DGGE), temperature gradient gel electrophoresis (TGGE), and flow cytometry, which are based on species-specific gene sequences, are available for use in diagnostic laboratories.     

Molecular phylogenetics of representative Paramecium species

The genus Paramecium has been known to science for 250 years and contains some of the most widely studied species of ciliates. At present, the basic research object for phylogenetic studies is the genome of various paramecia. One of the most widely used markers are genes coding for various rRNA's. Comparative analyses of sequences coding rRNA were applied for resolving the systematic position of some paramecia species and also for the establishment of an accurate taxonomy of Paramecium. Paramecia were also model organisms for their systematic group in more general studies in a comparative analysis among ciliates, fungi, plants and multicellular animals, illustrating the evolutionary relationships between Archaebacteria and Eucaryota. A new, revolutionary genealogy proposed the shifting of presumptively advanced groups towards more primitive ones, and traditionally primitive forms were located closer to highly specialized taxa, but rRNA analysis did not unambiguously resolve associations within the studied groups. Because of the aforementioned concerns, the number of molecular markers used for alternative studies is growing, such as genes coding proteins from the Hsp family or histone proteins. Other promising candidate markers may be hemoglobin genes or genes coding á-tubulins. In case of comparative analyses ofnucleotide sequences, the outcome of the research usually depends upon a subjective choice of DNA. One of the directions of research in molecular phylogenetics include indirect methods that allow for an estimation of entire genomes, for example RAPD-PCR-fingerprinting.     

Immunophilins and parvulins. Superfamily of peptidyl prolyl isomerases in Arabidopsis

Immunophilins are defined as receptors for immunosuppressive drugs including cyclosporin A, FK506, and rapamycin. The cyclosporin A receptors are referred to as cyclophilins (CYPs) and FK506- and rapamycin-binding proteins are abbreviated as FKBPs. These two groups of proteins (collectively called immunophilins) share little sequence homology, but both have peptidyl prolyl cis/trans isomerase (PPIase) activity that is involved in protein folding processes. Studies have identified immunophilins in all organisms examined including bacteria, fungi, animals, and plants. Nevertheless, the physiological function of immunophilins is poorly understood in any organism. In this study, we have surveyed the genes encoding immunophilins in Arabidopsis genome. A total of 52 genes have been found to encode putative immunophilins, among which 23 are putative FKBPs and 29 are putative CYPs. This is by far the largest immunophilin family identified in any organism. Both FKBPs and CYPs can be classified into single domain and multiple domain members. The single domain members contain a basic catalytic domain and some of them have signal sequences for targeting to a specific organelle. The multiple domain members contain not only the catalytic domain but also defined modules that are involved in protein-protein interaction or other functions. A striking feature of immunophilins in Arabidopsis is that a large fraction of FKBPs and CYPs are localized in the chloroplast, a possible explanation for why plants have a larger immunophilin family than animals. Parvulins represent another family of PPIases that are unrelated to immunophilins in protein sequences and drug binding properties. Three parvulin genes were found in Arabidopsis genome. The expression of many immunophilin and parvulin genes is ubiquitous except for those encoding chloroplast members that are often detected only in the green tissues. The large number of genes and diversity of structure domains and cellular localization make PPIases a versatile superfamily of proteins that clearly function in many cellular processes in plants.     

Protein structures and split genes

Exon-intron structures of eukaryotic genes were examined closely in their relation to primary and tertiary structures of the proteins they encode. Specific attention was given to the introns of genes encoding proteins having no repeats in their amino acid sequences. such introns have been shown to be located at sites corresponding to inter-domain or inter-module junctions of proteins identified in their three dimensional structures. “Modules,” compact structural units in globular domains of proteins, are identified by drawing a distance map. Intron positions are found to correspond to inter-module junctions in various proteins whose X-ray crystallographic data are available: the glogin family, CEWL, ovomucoid, cytochrome c, ADH, and trypsin-like serine proteinases.The good correspondence between intron positions and inter-module junctions excludes a mechanism of random insertion of introns, because the probability of intron insertion at each inter-module junction is extraordinarily small. Intron positions have been very stable and well conserved during evolution. However, at some inter-module junctions no introns are found.Modules in small proteins having no core modules buried in their interior have a character suitable for recruitment through their assembly into a stable domain; one side of them is rich in hydrophobic residues and the other in hydrophilic residues. Functionally important residues are scattered on different modules in the proteins examined. Based on these observations, the role of modules in the precellular period was conjectured: some of them might be functionally active by themselves but most modules might be only segments who could functions as an active protein only in an assembly. The origin of introns might be traced back prior to the divergence of prokaryotes and eukaryotes.     

A screen of cell-surface molecules identifies leucine-rich repeat proteins as key mediators of synaptic target selection

In Drosophila embryos and larvae, a small number of identified motor neurons innervate body wall muscles in a highly stereotyped pattern. Although genetic screens have identified many proteins that are required for axon guidance and synaptogenesis in this system, little is known about the mechanisms by which muscle fibers are defined as targets for specific motor axons. To identify potential target labels, we screened 410 genes encoding cell-surface and secreted proteins, searching for those whose overexpression on all muscle fibers causes motor axons to make targeting errors. Thirty such genes were identified, and a number of these were members of a large gene family encoding proteins whose extracellular domains contain leucine-rich repeat (LRR) sequences, which are protein interaction modules. By manipulating gene expression in muscle 12, we showed that four LRR proteins participate in the selection of this muscle as the appropriate synaptic target for the RP5 motor neuron.     

Two highly thermostable paralogous single-stranded DNA-binding proteins from <Emphasis Type="Italic">Thermoanaerobacter tengcongensis</Emphasis>

The thermophilic bacterium Thermoanaerobacter tengcongensis has two single-stranded DNA-binding (SSB) proteins, designated TteSSB2 and TteSSB3. In a SSB complementation assay in Escherichia coli, only TteSSB3 took over the in vivo function of EcoSSB. We have cloned the ssb genes obtained by PCR and have developed E. coli overexpression systems. The TteSSB2 and TteSSB3 consist of 153 and 150 amino acids with a calculated molecular mass of 17.29 and 16.96 kDa, respectively. They are the smallest known bacterial SSB proteins. The homology between amino acid sequences of these proteins is 40% identity and 53% similarity. They are functional as homotetramers, with each monomer encoding one single-stranded DNA binding domain (OB-fold). In fluorescence titrations with poly(dT), both proteins bind single-stranded DNA with a binding site size of about 40 nt per homotetramer. Thermostability with half-life of about 30 s at 95 degrees C makes TteSSB3 similar to the known SSB of Thermus aquaticus (TaqSSB). The TteSSB2 was fully active even after 6 h incubation at 100 degrees C. Here, we show for the first time paralogous thermostable homotetrameric SSBs, which could be an attractive alternative for known homodimeric thermostable SSB proteins in their applications for molecular biology methods and analytical purposes.     

Eukaryotic expression vectors containing genes encoding plant proteins for killing of cancer cells

Background: Gene therapy has attracted attention for its potential to specifically and efficiently target cancer cells with minimal toxicity to normal cells. At present, it offers a promising direction for the treatment of cancer patients. Numerous vectors have been engineered for the sole purpose of killing cancer cells, and some have successfully suppressed malignant tumours. Many plant proteins have anticancer properties; consequently, genes encoding some of these proteins are being used to design constructs for the inhibition of multiplying cancer cells. Results: Data addressing the function of vectors harbouring genes specifically encoding ricin, saporin, lunasin, linamarase, and tomato thymidine kinase 1 under the control of different promoters are summarised here. Constructs employing genes to encode cytotoxic proteins as well as constructs employing genes of enzymes that convert a nontoxic prodrug into a toxic drug are considered here. Conclusion: Generation of eukaryotic expression vectors containing genes encoding plant proteins for killing of cancer cells may permit the broadening of cancer gene therapy strategy, particularly because of the specific mode of action of anticancer plant proteins.     

用RACE结合cDNA文库筛选的方法获取新的锌指蛋白基因

大多数有重要功能的蛋白质都含相应的由保守氨基酸顺序组成的功能结构域。本文首先根据蛋白质功能结构域保守氨基酸序列设计简并引物,用PCR方法扩增出基因EST序列,再利用改进的快速扩增cDNA末端（RACE）方法从cDNA文库中扩增出基因非同源部位,然后以非同源序列为探针,筛选cDNA文库。利用此方法成功地从人骨髓cDNA文库中克隆到几个编码锌指蛋白并代表原有EST的新的全长cDNA。这一策略也应适用于筛选编码具有其他序列保守性功能结构域蛋白的基因。 Abstract:Most of the important functionally proteins contain the corresponding function domains that consist of conserved amino acid sequences.The study provided a method to identify novel genes that encode proteins containing important functionally domains with conserved sequences.First,primers were designed according to the sequence of the cDNA library vector and the ESTs that have been obtained by reverse PCR and degenerate primers encoding Zinc finger domain.The cDNA library DNA was used as template for PCR amplification.The amplified fragment that contains nonhomologous sequences of the cDNA was inserted into pGEM-T easy vector.The fragment was recovered and used as a probe for screening the cDNA library.Several cDNAs with full length that encode proteins with Zinc finger domain and represent the original ESTs have been successfully cloned from a human bone marrow cDNA library.This strategy can also be used in screening genes that encode proteins containing differential function domains with conserved sequences.     

Metagenomic insights into the fibrolytic microbiome in yak rumen

The rumen hosts one of the most efficient microbial systems for degrading plant cell walls, yet the predominant cellulolytic proteins and fibrolytic mechanism(s) remain elusive. Here we investigated the cellulolytic microbiome of the yak rumen by using a combination of metagenome-based and bacterial artificial chromosome (BAC)-based functional screening approaches. Totally 223 fibrolytic BAC clones were pyrosequenced and 10,070 ORFs were identified. Among them 150 were annotated as the glycoside hydrolase (GH) genes for fibrolytic proteins, and the majority (69%) of them were clustered or linked with genes encoding related functions. Among the 35 fibrolytic contigs of >10 Kb in length, 25 were derived from Bacteroidetes and four from Firmicutes. Coverage analysis indicated that the fibrolytic genes on most Bacteroidetes-contigs were abundantly represented in the metagenomic sequences, and they were frequently linked with genes encoding SusC/SusD-type outer-membrane proteins. GH5, GH9, and GH10 cellulase/hemicellulase genes were predominant, but no GH48 exocellulase gene was found. Most (85%) of the cellulase and hemicellulase proteins possessed a signal peptide; only a few carried carbohydrate-binding modules, and no cellulosomal domains were detected. These findings suggest that the SucC/SucD-involving mechanism, instead of one based on cellulosomes or the free-enzyme system, serves a major role in lignocellulose degradation in yak rumen. Genes encoding an endoglucanase of a novel GH5 subfamily occurred frequently in the metagenome, and the recombinant proteins encoded by the genes displayed moderate Avicelase in addition to endoglucanase activities, suggesting their important contribution to lignocellulose degradation in the exocellulase-scarce rumen.     

Recombination and chimeragenesis by in vitro heteroduplex formation and in vivo repair.

We describe a simple method for creating libraries of chimeric DNA sequences derived from homologous parental sequences. A heteroduplex formed in vitro is used to transform bacterial cells where repair of regions of non-identity in the heteroduplex creates a library of new, recombined sequences composed of elements from each parent. Heteroduplex recombination provides a convenient addition to existing DNA recombination methods ('DNA shuffling') and should be particularly useful for recombining large genes or entire operons. This method can be used to create libraries of chimeric polynucleotides and proteins for directed evolution to improve their properties or to study structure-function relationships. We also describe a simple test system for evaluating the performance of DNA recombination methods in which recombination of genes encoding truncated green fluorescent protein (GFP) reconstructs the full-length gene and restores its characteristic fluorescence. Comprising seven truncated GFP constructs, this system can be used to evaluate the efficiency of recombination between mismatches separated by as few as 24 bp and as many as 463 bp. The optimized heteroduplex recombination protocol is quite efficient, generating nearly 30% fluorescent colonies for recombination between two genes containing stop codons 463 bp apart (compared to a theoretical limit of 50%).     

Expression of human plasma protein genes in ageing transgenic mice

Introduction of human plasma protein genes into the mouse genome to produce transgenic mice furnishes an in vivo model for correlating chromosomal DNA sequences with developmental and tissue-specific expression. The liver produces an array of plasma proteins that circulate throughout the body contributing to homeostasis. Non-hepatic tissue sites of synthesis have been identified where a local provision of plasma proteins in needed. Analysis of expression of human plasma protein genes in ageing transgenic mice appears especialy promising in identifying DNA sequences that respond to environmental adversities such as inflammatory factors, hormonal changes and metal toxicity. The results indicate that human genes encoding and controlling liver plasma proteins serve as useful models for studying genetic regulation in the background of development and ageing.     

Identification of putative parasitism genes expressed in the esophageal gland cells of the soybean cyst nematode Heterodera glycines

Cloning parasitism genes encoding secretory proteins expressed in the esophageal gland cells is the key to understanding the molecular basis of nematode parasitism of plants. Suppression subtractive hybridization (SSH) with the microaspirated contents from Heterodera glycines esophageal gland cells and intestinal region was used to isolate genes expressed preferentially in the gland cells of parasitic stages. Twenty-three unique cDNA sequences from a SSH cDNA library were identified and hybridized to the genomic DNA of H. glycines in Southern blots. Full-length cDNAs of 21 clones were obtained by screening a gland-cell long-distance polymerase chain reaction cDNA library. Deduced proteins of 10 clones were preceded by a signal peptide for secretion, and PSORT II computer analysis predicted eight proteins as extracellular, one as nuclear, and one as plasmalemma localized. In situ hybridization showed that four of the predicted extracellular clones were expressed specifically in the dorsal gland cell, one in the subventral gland cells, and three in the intestine in H. glycines. The predicted nuclear clone and the plasmalemma-localized clone were expressed in the subventral gland cells and the dorsal gland cell, respectively. SSH is an efficient method for cloning putative parasitism genes encoding esophageal gland cell secretory proteins that may have a role in H. glycines parasitism of soybean.     

Computer-simulation-based selection of optimal monomer for imprinting of tri-O-acetyl adenosine in a polymer matrix: calculations for benzene solution

Molecular imprinting is a promising way to create polymer materials that can be used as artificial receptors, and have anticipated use in synthetic imitation of natural antibodies. In case of successful imprinting, the selectivity and affinity of the imprint for the substrate molecules are comparable with those of natural counterparts. Various calculation methods can be used to estimate the effects of a large range of imprinting parameters under different conditions, and to find better ways to improve polymer characteristics. However, one difficulty is that properties such as hydrogen bonding can be modeled only by quantum methods that demand a lot of computational resources. Combined quantum mechanics/molecular mechanics (QM/MM) methods allow the use of MM and QM for different parts of the modeled system. In present study this method was realized in the NWChem package to compare estimations of the stability of tri-O-acetyl adenosine–monomer pre-polymerization complexes in benzene solution with previous results under vacuum.     

Assessing the biological significance of gene expression signatures and co-expression modules by studying their network properties

Microarray experiments have been extensively used to define signatures, which are sets of genes that can be considered markers of experimental conditions (typically diseases). Paradoxically, in spite of the apparent functional role that might be attributed to such gene sets, signatures do not seem to be reproducible across experiments. Given the close relationship between function and protein interaction, network properties can be used to study to what extent signatures are composed of genes whose resulting proteins show a considerable level of interaction (and consequently a putative common functional role).We have analysed 618 signatures and 507 modules of co-expression in cancer looking for significant values of four main protein-protein interaction (PPI) network parameters: connection degree, cluster coefficient, betweenness and number of components. A total of 3904 gene ontology (GO) modules, 146 KEGG pathways, and 263 Biocarta pathways have been used as functional modules of reference.Co-expression modules found in microarray experiments display a high level of connectivity, similar to the one shown by conventional modules based on functional definitions (GO, KEGG and Biocarta). A general observation for all the classes studied is that the networks formed by the modules improve their topological parameters when an external protein is allowed to be introduced within the paths (up to the 70% of GO modules show network parameters beyond the random expectation). This fact suggests that functional definitions are incomplete and some genes might still be missing. Conversely, signatures are clearly not capturing the altered functions in the corresponding studies. This is probably because the way in which the genes have been selected in the signatures is too conservative. These results suggest that gene selection methods which take into account relationships among genes should be superior to methods that assume independence among genes outside their functional contexts.     

MDC9, a widely expressed cellular disintegrin containing cytoplasmic SH3 ligand domains

Cellular disintegrins are a family of proteins that are related to snake venom integrin ligands and metalloproteases. We have cloned and sequenced the mouse and human homologue of a widely expressed cellular disintegrin, which we have termed MDC9 (for metalloprotease/disintegrin/cysteine-rich protein 9). The deduced mouse and human protein sequences are 82% identical. MDC9 contains several distinct protein domains: a signal sequence is followed by a prodomain and a domain with sequence similarity to snake venom metalloproteases, a disintegrin domain, a cysteine-rich region, an EGF repeat, a membrane anchor, and a cytoplasmic tail. The cytoplasmic tail of MDC9 has two proline-rich sequences which can bind the SH3 domain of Src, and may therefore function as SH3 ligand domains. Western blot analysis shows that MDC9 is an approximately 84-kD glycoprotein in all mouse tissues examined, and in NIH 3T3 fibroblast and C2C12 myoblast mouse cell lines. MDC9 can be both cell surface biotinylated and 125I-labeled in NIH 3T3 mouse fibroblasts, indicating that the protein is present on the plasma membrane. Expression of MDC9 in COS-7 cells yields an 84-kD protein, and immunofluorescence analysis of COS-7 cells expressing MDC9 shows a staining pattern that is consistent with a plasma membrane localization. The apparent molecular mass of 84 kD suggests that MDC9 contains a membrane-anchored metalloprotease and disintegrin domain. We propose that MDC9 might function as a membrane-anchored integrin ligand or metalloprotease, or that MDC9 may combine both activities in one protein.     

A transmembrane trap method for efficient cloning of genes encoding proteins possessing transmembrane domain.

To facilitate searching for genes encoding cell membrane proteins, we developed a method for isolating cDNAs that contain sequences for hydrophobic transmembrane runs. This cloning strategy, termed the "transmembrane (TM) trap method," utilizes a vector that directs the cell surface expression of mouse CD4 fusion protein when an insert encoding hydrophobic transmembrane sequences is cloned in-frame with correct orientation. We applied this novel method to isolation of cytokine receptor cDNAs. Our strategy enabled efficient isolation of relatively rare species encoding receptors such as IL-2Rgamma, IL-3Rbeta, IL-4Ralpha, IL-5Ralpha, and IL-6Ralpha. This method also could be used to isolate cDNAs for intracellular molecules with a transmembrane region, e.g., bcl-2. These results indicate that the TM trap method provides an efficient cloning strategy for identification of various families of genes encoding proteins with one or more transmembrane regions.