Inference of gene pathways using mixture Bayesian networks

Background  

Inference of gene networks typically relies on measurements across a wide range of conditions or treatments. Although one network structure is predicted, the relationship between genes could vary across conditions. A comprehensive approach to infer general and condition-dependent gene networks was evaluated. This approach integrated Bayesian network and Gaussian mixture models to describe continuous microarray gene expression measurements, and three gene networks were predicted.     

Bayesian Inference on Proportional Elections

Polls for majoritarian voting systems usually show estimates of the percentage of votes for each candidate. However, proportional vote systems do not necessarily guarantee the candidate with the most percentage of votes will be elected. Thus, traditional methods used in majoritarian elections cannot be applied on proportional elections. In this context, the purpose of this paper was to perform a Bayesian inference on proportional elections considering the Brazilian system of seats distribution. More specifically, a methodology to answer the probability that a given party will have representation on the chamber of deputies was developed. Inferences were made on a Bayesian scenario using the Monte Carlo simulation technique, and the developed methodology was applied on data from the Brazilian elections for Members of the Legislative Assembly and Federal Chamber of Deputies in 2010. A performance rate was also presented to evaluate the efficiency of the methodology. Calculations and simulations were carried out using the free R statistical software.     

Expression of the Arabidopsis thaliana invertase gene family

Two new loci have been found to be clustered with five other genes for the nitrate assimilation pathway in the Chlamydomonas reinhardtii genome. One gene, located close to the 3′-end of the high-affinity nitrate transporter (HANT) gene Nrt2;2, corresponds to the nitrite reductase (NiR) structural gene Nii1. This is supported by a number of experimental findings: (i) NiR-deficient mutants have lost Nii1 gene expression; (ii) Nii1 mRNA accumulation is co-regulated with the expression of other structural genes of the nitrate assimilation pathway; (iii) nitrite (nitrate) utilization ability is recovered in the NiR mutants by functional complementation with a wild-type Nii1 gene; (iv) the elucidated NII1 amino acid sequence is highly similar to that of the cyanobacterial and higher-plant enzyme, and contains the predicted domains for plastidic ferredoxin-NiRs. Thus, the mutant phenotype and the mRNA sequence and expression of the Nii1 gene have been unequivocally related. Accumulation of mRNA for the second locus identified, Lde1 (light-dependent expression), was not regulated by nitrogen, but like nitrate-assimilation clustered genes, its expression was down-regulated in the dark. Received: 27 November 1997 / Accepted: 19 January 1998     

The Arabidopsis acyl-CoA oxidase gene family

A family of acyl-CoA oxidase isozymes catalyse the first step in the peroxisomal fatty acid beta-oxidation spiral. Our group and others have recently characterized four genes from this family in the model oilseed Arabidopsis. These genes encode isozymes with different acyl-CoA substrate specificities, which together encompass the full range of fatty acid chain lengths that exist in vivo. Here we review the biochemical properties and physiological roles of the acyl-CoA oxidase isozymes.     

Functional divergence in the Arabidopsis LOB-domain gene family

The Arabidopsis LOB-domain (LBD) gene family is composed by 43 members divided in two classes based on amino acid conservation within the LOB-domain. The LOB domain is known to be responsible for DNA binding and protein-protein interactions. There is very little functional information available for most genes in the LBD family and many lbd single mutants do not exhibit conspicuous phenotypes. One plausible explanation for the limited loss-of-function phenotypes observed in this family is that LBD genes exhibit significant functional redundancy. Here we discuss an example of one phylogenetic subgroup of the LBD family, in which genes that are closely related based on phylogeny exhibit distinctly different expression patterns and do not have overlapping functions. We discuss the challenges of using phylogenetic analyses to predict redundancy in gene families.     

Analysis of the sucrose synthase gene family in Arabidopsis

The properties and expression patterns of the six isoforms of sucrose synthase in Arabidopsis are described, and their functions are explored through analysis of T-DNA insertion mutants. The isoforms have generally similar kinetic properties. Although there is variation in sensitivity to substrate inhibition by fructose this is unlikely to be of major physiological significance. No two isoforms have the same spatial and temporal expression patterns. Some are highly expressed in specific locations, whereas others are more generally expressed. More than one isoform is expressed in all organs examined. Mutant plants lacking individual isoforms have no obvious growth phenotypes, and are not significantly different from wild-type plants in starch, sugar and cellulose content, seed weight or seed composition under the growth conditions employed. Double mutants lacking the pairs of similar isoforms sus2 and sus3, and sus5 and sus6, are also not significantly different in these respects from wild-type plants. These results are surprising in the light of the marked phenotypes observed when individual isoforms are eliminated in crop plants including pea, maize, potato and cotton. A sus1/sus4 double mutant grows normally in well-aerated conditions, but shows marked growth retardation and accumulation of sugars when roots are subjected to hypoxia. The sucrose synthase activity in roots of this mutant is 3% or less of wild-type activity. Thus under well-aerated conditions sucrose mobilization in the root can proceed almost entirely via invertases without obvious detriment to the plant, but under hypoxia there is a specific requirement for sucrose synthase activity.     

The PRA1 gene family in Arabidopsis

Prenylated Rab acceptor 1 (PRA1) domain proteins are small transmembrane proteins that regulate vesicle trafficking as receptors of Rab GTPases and the vacuolar soluble N-ethylmaleimide-sensitive factor attachment receptor protein VAMP2. However, little is known about PRA1 family members in plants. Sequence analysis revealed that higher plants, compared with animals and primitive plants, possess an expanded family of PRA1 domain-containing proteins. The Arabidopsis (Arabidopsis thaliana) PRA1 (AtPRA1) proteins were found to homodimerize and heterodimerize in a manner corresponding to their phylogenetic distribution. Different AtPRA1 family members displayed distinct expression patterns, with a preference for vascular cells and expanding or developing tissues. AtPRA1 genes were significantly coexpressed with Rab GTPases and genes encoding vesicle transport proteins, suggesting an involvement in the vesicle trafficking process similar to that of their animal counterparts. Correspondingly, AtPRA1 proteins were localized in the endoplasmic reticulum, Golgi apparatus, and endosomes/prevacuolar compartments, hinting at a function in both secretory and endocytic intracellular trafficking pathways. Taken together, our data reveal a high functional diversity of AtPRA1 proteins, probably dealing with the various demands of the complex trafficking system.     

Functional diversity inside the Arabidopsis polyamine oxidase gene family

Polyamine oxidases (PAOs) are FAD-dependent enzymes involved in polyamine catabolism. All so far characterized PAOs from monocotyledonous plants, such as the apoplastic maize PAO, oxidize spermine (Spm) and spermidine (Spd) to produce 1,3-diaminopropane, H(2)O(2), and an aminoaldehyde, and are thus considered to be involved in a terminal catabolic pathway. Mammalian PAOs oxidize Spm or Spd (and/or their acetyl derivatives) differently from monocotyledonous PAOs, producing Spd or putrescine, respectively, in addition to H(2)O(2) and an aminoaldehyde, and are therefore involved in a polyamine back-conversion pathway. In Arabidopsis thaliana, five PAOs (AtPAO1-AtPAO5) are present with cytosolic or peroxisomal localization and three of them (the peroxisomal AtPAO2, AtPAO3, and AtPAO4) form a distinct PAO subfamily. Here, a comparative study of the catalytic properties of recombinant AtPAO1, AtPAO2, AtPAO3, and AtPAO4 is presented, which shows that all four enzymes strongly resemble their mammalian counterparts, being able to oxidize the common polyamines Spd and/or Spm through a polyamine back-conversion pathway. The existence of this pathway in Arabidopsis plants is also evidenced in vivo. These enzymes are also able to oxidize the naturally occurring uncommon polyamines norspermine and thermospermine, the latter being involved in important plant developmental processes. Furthermore, data herein reveal some important differences in substrate specificity among the various AtPAOs, which suggest functional diversity inside the AtPAO gene family. These results represent a new starting point for further understanding of the physiological role(s) of the polyamine catabolic pathways in plants.     

A molecular-genetic study of the Arabidopsis Toc75 gene family

Toc75 (translocon at the outer envelope membrane of chloroplasts, 75 kD) is the protein translocation channel at the outer envelope membrane of plastids and was first identified in pea (Pisum sativum) using biochemical approaches. The Arabidopsis (Arabidopsis thaliana) genome contains three Toc75-related sequences, termed atTOC75-I, atTOC75-III, and atTOC75-IV, which we studied using a range of molecular, genetic, and biochemical techniques. Expression of atTOC75-III is strongly regulated and at its highest level in young, rapidly expanding tissues. By contrast, atTOC75-IV is expressed uniformly throughout development and at a much lower level than atTOC75-III. The third sequence, atTOC75-I, is a pseudogene that is not expressed due to a gypsy/Ty3 transposon insertion in exon 1, and numerous nonsense, frame-shift, and splice-junction mutations. The expressed genes, atTOC75-III and atTOC75-IV, both encode integral envelope membrane proteins. Unlike atToc75-III, the smaller atToc75-IV protein is not processed upon targeting to the envelope, and its insertion does not require ATP at high concentrations. The atTOC75-III gene is essential for viability, since homozygous atToc75-III knockout mutants (termed toc75-III) could not be identified, and aborted seeds were observed at a frequency of approximately 25% in the siliques of self-pollinated toc75-III heterozygotes. Homozygous toc75-III embryos were found to abort at the two-cell stage. Homozygous atToc75-IV knockout plants (termed toc75-IV) displayed no obvious visible phenotypes. However, structural abnormalities were observed in the etioplasts of toc75-IV seedlings and atTOC75-IV overexpressing lines, and toc75-IV plants were less efficient at deetiolation than wild type. These results suggest some role for atToc75-IV during growth in the dark.     

Characterization of the GGPP synthase gene family in Arabidopsis thaliana

Geranylgeranyl diphosphate (GGPP) is a key precursor of various isoprenoids that have diverse functions in plant metabolism and development. The annotation of the Arabidopsis thaliana genome predicts 12 genes to encode geranylgeranyl diphosphate synthases (GGPPS). In this study we analyzed GGPPS activity as well as the subcellular localization and tissue-specific expression of the entire protein family in A. thaliana. GGPPS2 (At2g18620), GGPPS3 (At2g18640), GGPPS6 (At3g14530), GGPPS7 (At3g14550), GGPPS8 (At3g20160), GGPPS9 (At3g29430), GGPPS10 (At3g32040) and GGPPS11 (At4g36810) showed GGPPS activity in Escherichia coli, similar to activities reported earlier for GGPPS1 (At1g49530) and GGPPS4 (At2g23800) (Zhu et al. in Plant Cell Physiol 38(3):357–361, 1997a; Plant Mol Biol 35(3):331–341, b). GGPPS12 (At4g38460) did not produce GGPP in E. coli. Based on DNA sequence analysis we propose that GGPPS5 (At3g14510) is a pseudogene. GGPPS–GFP (green fluorescent protein) fusion proteins of the ten functional GGPP synthases localized to plastids, mitochondria and the endoplasmic reticulum, with the majority of the enzymes located in plastids. Gene expression analysis using quantitative real time-PCR, GGPPS promoter-GUS (β-glucuronidase) assays and publicly available microarray data revealed a differential spatio-temporal expression of GGPPS genes. The results suggest that plastids and mitochondria are key subcellular compartments for the synthesis of ubiquitous GGPP-derived isoprenoid species. GGPPS11 and GGPPS1 are the major isozymes responsible for their biosynthesis. All remaining paralogs, encoding six plastidial isozymes and two cytosolic isozymes, were expressed in specific tissues and/or at specific developmental stages, suggesting their role in developmentally regulated isoprenoid biosynthesis. Our results show that of the 12 predicted GGPPS encoded in the A. thaliana genome 10 are functional proteins that can synthesize GGPP. Their specific subcellular location and differential expression pattern suggest subfunctionalization in providing GGPP to specific tissues, developmental stages, or metabolic pathways.