Constrained clusters of gene expression profiles with pathological features

MOTIVATION: Gene expression profiles should be useful in distinguishing variations in disease, since they reflect accurately the status of cells. The primary clustering of gene expression reveals the genotypes that are responsible for the proximity of members within each cluster, while further clustering elucidates the pathological features of the individual members of each cluster. However, since the first clustering process and the second classification step, in which the features are associated with clusters, are performed independently, the initial set of clusters may omit genes that are associated with pathologically meaningful features. Therefore, it is important to devise a way of identifying gene expression clusters that are associated with pathological features. RESULTS: We present the novel technique of 'itemset constrained clustering' (IC-Clustering), which computes the optimal cluster that maximizes the interclass variance of gene expression between groups, which are divided according to the restriction that only divisions that can be expressed using common features are allowed. This constraint automatically labels each cluster with a set of pathological features which characterize that cluster. When applied to liver cancer datasets, IC-Clustering revealed informative gene expression clusters, which could be annotated with various pathological features, such as 'tumor' and 'man', or 'except tumor' and 'normal liver function'. In contrast, the k-means method overlooked these clusters.     

Molecular classification of liver cirrhosis in a rat model by proteomics and bioinformatics

Liver cirrhosis is a worldwide health problem. Reliable, noninvasive methods for early detection of liver cirrhosis are not available. Using a three-step approach, we classified sera from rats with liver cirrhosis following different treatment insults. The approach consisted of: (i) protein profiling using surface-enhanced laser desorption/ionization (SELDI) technology; (ii) selection of a statistically significant serum biomarker set using machine learning algorithms; and (iii) identification of selected serum biomarkers by peptide sequencing. We generated serum protein profiles from three groups of rats: (i) normal (n=8), (ii) thioacetamide-induced liver cirrhosis (n=22), and (iii) bile duct ligation-induced liver fibrosis (n=5) using a weak cation exchanger surface. Profiling data were further analyzed by a recursive support vector machine algorithm to select a panel of statistically significant biomarkers for class prediction. Sensitivity and specificity of classification using the selected protein marker set were higher than 92%. A consistently down-regulated 3495 Da protein in cirrhosis samples was one of the selected significant biomarkers. This 3495 Da protein was purified on-chip and trypsin digested. Further structural characterization of this biomarkers candidate was done by using cross-platform matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS) peptide mass fingerprinting (PMF) and matrix-assisted laser desorption/ionization time of flight/time of flight (MALDI-TOF/TOF) tandem mass spectrometry (MS/MS). Combined data from PMF and MS/MS spectra of two tryptic peptides suggested that this 3495 Da protein shared homology to a histidine-rich glycoprotein. These results demonstrated a novel approach to discovery of new biomarkers for early detection of liver cirrhosis and classification of liver diseases.     

EST-based analysis of gene expression in the porcine brain

Since pig is an important livestock species worldwide, its gene expression has been investigated intensively, but rarely in brain. In order to study gene expression profiles in the pig central nervous system, we sequenced and analyzed 43,122 highquality 5’ end expressed sequence tags (ESTs) from porcine cerebellum, cortex cerebrum, and brain stem cDNA libraries, involving several different prenatal and postnatal developmental stages. The initial ESTs were assembled into 16,101 clusters and compared to protein and nucleic acid databases in GenBank. Of these sequences, 30.6% clusters matched protein databases and represented function known sequences; 75.1% had significant hits to nucleic acid databases and partial represented known function; 73.3% matched known porcine ESTs; and 21.5% had no matches to any known sequences in GenBank. We used the categories defined by the Gene Ontology to survey gene expression in the porcine brain.     

Genes controlling seed dormancy and pre-harvest sprouting in a rice-wheat-barley comparison

Pre-harvest sprouting results in significant economic loss for the grain industry around the world. Lack of adequate seed dormancy is the major reason for pre-harvest sprouting in the field under wet weather conditions. Although this trait is governed by multiple genes it is also highly heritable. A major QTL controlling both pre-harvest sprouting and seed dormancy has been identified on the long arm of barley chromosome 5H, and it explains over 70% of the phenotypic variation. Comparative genomics approaches among barley, wheat and rice were used to identify candidate gene(s) controlling seed dormancy and hence one aspect of pre-harvest sprouting. The barley seed dormancy/pre-harvest sprouting QTL was located in a region that showed good synteny with the terminal end of the long arm of rice chromosome 3. The rice DNA sequences were annotated and a gene encoding GA20-oxidase was identified as a candidate gene controlling the seed dormancy/pre-harvest sprouting QTL on 5HL. This chromosomal region also shared synteny with the telomere region of wheat chromosome 4AL, but was located outside of the QTL reported for seed dormancy in wheat. The wheat chromosome 4AL QTL region for seed dormancy was syntenic to both rice chromosome 3 and 11. In both cases, corresponding QTLs for seed dormancy have been mapped in rice.C. Li and P. Ni contributed equally to this work     

Anaplasma phagocytophilum AnkA binds to granulocyte DNA and nuclear proteins

Human granulocytic anaplasmosis (HGA) is caused by the obligate intracellular bacterium Anaplasma phagocytophilum. The bacterium infects, survives, propagates in, and alters neutrophil phenotype, indicating unique survival mechanisms. AnkA is the only known A. phagocytophilum component that gains access beyond neutrophil vacuoles and is transported to the infected host cell nucleus. The ability of native and recombinant AnkA to bind DNA and nuclear proteins from host HL-60 cells was assessed by the use of immunoprecipitation after cis-diamminedichloroplatinum (cis-DDP) DNA-protein crosslinking, by probing uninfected HL-60 cell nuclear lysates for AnkA binding, and by recovery and sequence analysis of immunoprecipitated DNA. AnkA binds HL-60 cell DNA as well as nuclear proteins of approximately 86, 53 and 25 kDa, whereas recombinant A. phagocytophilum Msp2 or control proteins do not. DNA immunoprecipitation reveals AnkA binding to a variety of target genes in the human genome, including genes that encode proteins with ATPase, tyrosine phosphatase and NADH dehydrogenase-like functions. These data indicate that AnkA could exert some effect on cells through binding to protein:DNA complexes in neutrophil nuclei. Whether AnkA binding leads to neutrophil functional alterations, and how such alterations might occur will depend upon definitive identification of binding partners and associated metabolic and biochemical pathways.     

Recombinant antigens for immunodiagnosis of cystic echinococcosis

Three cDNAs, termed EpC1, TPxEg and EgG5, were isolated by immunoscreening from an Echinococcus granulosus cDNA library. The recombinant phages exhibited strong reactivity with sera from humans with confirmed cystic echinococcosis (CE) and with sera from mice infected with E. granulosus oncospheres. The cDNAs were subcloned into a pET vector, expressed as fusion proteins tagged with GST and affinity purified against the GST tag. Of the three recombinant proteins, EpC1 achieved the highest performance for serodiagnosis of CE in Western blot analysis using a panel of clinically defined human sera to initially address the sensitivity and specificity of the molecules. The protein yielded an overall sensitivity of 92.2% and specificity of 95.6%, levels unprecedented taking into account the large panel of 896 human sera that were tested. The strategy used may also prove suitable for improved immunodiagnosis of other parasitic infections.     

Uniaxial drawing and mechanical properties of poly[(R)-3-hydroxybutyrate]/poly(L-lactic acid) blends

Blends of poly(L-lactic acid) (PLLA) with two kinds of poly[(R)-3-hydroxybutyrate] (PHB) having different molecular weights, commercial-grade bacterial PHB (bacterial-PHB) and ultrahigh molecular weight PHB (UHMW-PHB), were prepared by the solvent-casting method and uniaxially drawn at two drawing temperatures, around PHB's T(g) (2 degrees C) for PHB-rich blends and around PLLA's T(g) (60 degrees C) for PLLA-rich blends. Differential scanning calorimetry analysis showed that this system was immiscible over the entire composition range. Mechanical properties of all of the samples were improved in proportion to the draw ratio. Although PLLA domains in bacterial-PHB-rich blends remained almost unstretched during cold drawing, a good interfacial adhesion between two polymers and the reinforcing role of PLLA components led to enhanced mechanical properties proportionally to the PLLA content at the same draw ratio. On the contrary, in the case of UHMW-PHB-rich blends, the minor component PLLA was found to be also oriented by cold drawing in ice water due to an increase in the interfacial entanglements caused by the very long chain length of the matrix polymer. As a result, their mechanical properties were considerably improved with increasing PLLA content compared with the bacterial-PHB system. Scanning electron microscopy observations on the surface and cross-section revealed that a layered structure with uniformly oriented microporous in the interior was obtained by selectively removal of PLLA component after simple alkaline treatment.     

Interaction with IQGAP1 links APC to Rac1, Cdc42, and actin filaments during cell polarization and migration

Rho family GTPases, particularly Rac1 and Cdc42, are key regulators of cell polarization and directional migration. Adenomatous polyposis coli (APC) is also thought to play a pivotal role in polarized cell migration. We have found that IQGAP1, an effector of Rac1 and Cdc42, interacts directly with APC. IQGAP1 and APC localize interdependently to the leading edge in migrating Vero cells, and activated Rac1/Cdc42 form a ternary complex with IQGAP1 and APC. Depletion of either IQGAP1 or APC inhibits actin meshwork formation and polarized migration. Depletion of IQGAP1 or APC also disrupts localization of CLIP-170, a microtubule-stabilizing protein that interacts with IQGAP1. Taken together, these results suggest a model in which activation of Rac1 and Cdc42 in response to migration signals leads to recruitment of IQGAP1 and APC which, together with CLIP-170, form a complex that links the actin cytoskeleton and microtubule dynamics during cell polarization and directional migration.     

Solution structure of the RWD domain of the mouse GCN2 protein

GCN2 is the alpha-subunit of the only translation initiation factor (eIF2alpha) kinase that appears in all eukaryotes. Its function requires an interaction with GCN1 via the domain at its N-terminus, which is termed the RWD domain after three major RWD-containing proteins: RING finger-containing proteins, WD-repeat-containing proteins, and yeast DEAD (DEXD)-like helicases. In this study, we determined the solution structure of the mouse GCN2 RWD domain using NMR spectroscopy. The structure forms an alpha + beta sandwich fold consisting of two layers: a four-stranded antiparallel beta-sheet, and three side-by-side alpha-helices, with an alphabetabetabetabetaalphaalpha topology. A characteristic YPXXXP motif, which always occurs in RWD domains, forms a stable loop including three consecutive beta-turns that overlap with each other by two residues (triple beta-turn). As putative binding sites with GCN1, a structure-based alignment allowed the identification of several surface residues in alpha-helix 3 that are characteristic of the GCN2 RWD domains. Despite the apparent absence of sequence similarity, the RWD structure significantly resembles that of ubiquitin-conjugating enzymes (E2s), with most of the structural differences in the region connecting beta-strand 4 and alpha-helix 3. The structural architecture, including the triple beta-turn, is fundamentally common among various RWD domains and E2s, but most of the surface residues on the structure vary. Thus, it appears that the RWD domain is a novel structural domain for protein-binding that plays specific roles in individual RWD-containing proteins.