Weiwen Zhang

Integrative Analysis of Transcriptomic and Proteomic Data: Challenges, Solutions and Applications

ABSTRACT Recent advances in high-throughput technologies enable quantitative monitoring of the abundance of various biological molecules and allow determination of their variation between biological states on a genomic scale. Two popular platforms are DNA microarrays that measure messenger RNA transcript levels, and gel-free proteomic analyses that quantify protein abundance. Obviously, no single approach can fully unravel the complexities of fundamental biology and it is equally clear that integrative analysis of multiple levels of gene expression would be valuable in this endeavor. However, most integrative transcriptomic and proteomic studies have thus far either failed to find a correlation or only observed a weak correlation. In addition to various biological factors, it is suggested that the poor correlation could be quite possibly due to the inadequacy of available statistical tools to compensate for biases in the data collection methodologies. To address this issue, attempts have recently been made to systematically investigate the correlation patterns between transcriptomic and proteomic datasets, and to develop sophisticated statistical tools to improve the chances of capturing a relationship. The goal of these efforts is to enhance understanding of the relationship between transcriptomes and proteomes so that integrative analyses may be utilized to reveal new biological insights that are not accessible through one-dimensional datasets. In this review, we outline some of the challenges associated with integrative analyses and present some preliminary statistical solutions. In addition, some new applications of integrated transcriptomic and proteomic analysis to the investigation of post-transcriptional regulation are also discussed.

Global transcriptomic analysis of Desulfovibrio vulgaris on different electron donors.

Whole-genome microarrays of Desulfovibrio vulgaris were used to determine relative transcript levels in cells grown to exponential or stationary phase on a medium containing either lactate or formate as electron donor. The results showed that 158 and 477 genes were differentially expressed when comparing exponential to stationary phase in lactate- or formate-based media, respectively; and 505 and 355 genes were responsive to the electron donor used at exponential or stationary phase, respectively. Functional analyses suggested that the differentially regulated genes were involved in almost every aspect of cellular metabolism, with genes involved in protein synthesis, carbon, and energy metabolism being the most regulated. The results suggested that HynBA-1 might function as a primary periplasmic hydrogenase responsible for oxidation of H2 linked to the proton gradient in lactate-based medium, while several periplasmic hydrogenases including HynBA-1 and Hyd might carry out this role in formate-based medium. The results also indicated that the alcohol dehydrogenase and heterodisulfide reductase catalyzed pathway for proton gradient formation might be actively functioning for ATP synthesis in D. vulgaris. In addition, hierarchical clustering analysis using expression data across different electron donors and growth phases allowed the identification of the common electron donor independent changes in gene expression specifically associated with the exponential to stationary phase transition, and those specifically associated with the different electron donors independent of growth phase. The study provides the first global description and functional interpretation of transcriptomic response to growth phase and electron donor in D. vulgaris.

Oxidative stress and heat-shock responses in Desulfovibrio vulgaris by genome-wide transcriptomic analysis

Sulfate-reducing bacteria such as Desulfovibrio vulgaris have developed a set of responses that allow them to survive in hostile environments. To obtain further knowledge of the protective mechanisms employed by D.␣vulgaris in response to oxidative stress and heat shock, we performed a genome-wide transcriptomic analysis to determine the cellular responses to both stimuli. The results showed that 130 genes were responsive to oxidative stress, while 427 genes were responsive to heat-shock. Functional analyses suggested that the genes regulated were involved in a variety of cellular functions. Amino acid biosynthetic pathways were induced by both oxidative stress and heat shock treatments, while fatty acid metabolism, purine and cofactor biosynthesis were induced by heat shock only. The rubrerythrin gene (rbr) was up-regulated in response to oxidative stress, suggesting an important role for this protein in the oxidative damage resistance response in D. vulgaris. In addition, thioredoxin reductase (trxB) was also responsive to oxidative stress, suggesting that the thiol-specific redox system might also be involved in oxidative protection in this organism. In contrast, the expression of rubredoxin oxidoreductase (rbo), superoxide dismutase (sodB) and catalase (katA) genes were not regulated in response to oxidative stress. Comparison of cellular responses to oxidative stress and heat-shock allowed the identification of 66 genes that showed a similar drastic response to both environmental perturbations, implying that these genes might be part of the general stress response (GSR) network in D. vulgaris. This hypothesis was further supported by the identification of a conserved motif upstream of these stress-responsive genes.

Integrated analysis of transcriptomic and proteomic data of Desulfovibrio vulgaris: zero-inflated Poisson regression models to predict abundance of undetected proteins

MOTIVATIONnIntegrated analysis of global scale transcriptomic and proteomic data can provide important insights into the metabolic mechanisms underlying complex biological systems. However, because the relationship between protein abundance and mRNA expression level is complicated by many cellular and physical processes, sophisticated statistical models need to be developed to capture their relationship.nnnRESULTSnIn this study, we describe a novel data-driven statistical model to integrate whole-genome microarray and proteomic data collected from Desulfovibrio vulgaris grown under three different conditions. Based on the Poisson distribution pattern of proteomic data and the fact that a large number of proteins were undetected (excess zeros), zero-inflated Poisson (ZIP)-based models were proposed to define the correlation pattern between mRNA and protein abundance. In addition, by assuming that there is a probability mass at zero representing unexpressed genes and expressed proteins that were undetected owing to technical limitations, a Potential ZIP model was established. Two significant improvements introduced by this approach are (1) the predicted protein abundance level values for experimentally detected proteins are corrected by considering their mRNA levels and (2) protein abundance values can be predicted for undetected proteins (in the case of this study, approximately 83% of the proteins in the D.vulgaris genome) for better biological interpretation. We demonstrated the use of these statistical models by comparatively analyzing proteomic and microarray results from D.vulgaris grown on lactate-based versus formate-based media. These models correctly predicted increased expression of Ech hydrogenase and decreased expression of Coo hydrogenase for D.vulgaris grown on formate.

Evolution of the PPM-family protein phosphatases in Streptomyces: duplication of catalytic domain and lateral recruitment of additional sensory domains

Originally identified from eukaryotes, the Mg2+- or Mn2+-dependent protein phosphatases (PPMs) are a diverse group of enzymes whose members include eukaryotic PP2C and some prokaryotic serine/threonine phosphatases. In a previous study, unexpectedly large numbers of PPMs were identified in two Streptomyces genomes. In this work, a phylogenetic analysis was performed with all the PPMs available from a wide variety of microbial sources to determine the evolutionary origin of the Streptomyces PPM proteins. Consistent with earlier hypotheses, the results suggested that the microbial PPMs were relatively recent additions from eukaryotic sources. Results also indicated that the Streptomyces PPMs were divided into two major subfamilies at an early stage of their emergence in Streptomyces genomes. The first subfamily, which contains only six Streptomyces PPMs, possesses a catalytic domain whose sequence and architecture are similar to that of eukaryotic PPMs; the second subfamily contains 89 Streptomyces PPMs that lack the 5a and 5b catalytic domain motifs, similar to the PPMs SpoIIE and RsbU of Bacillus subtilis. Significant gene duplication was observed for the PPMs in the second subfamily. In addition, more than half (54 %) of the Streptomyces PPMs from the second subfamily were found to have at least one additional sensory domain, most commonly the PAS or the GAF domain. Phylogenetic analysis showed that these domains tended to be clustered according to the putative physiological functions rather than taxonomic relationship, implying that they might have arisen as a result of domain recruitment in a late evolutionary stage. This study provides an insight into how Streptomyces spp. may have expanded their PPM-based signal transduction networks to enable them to respond to a greater range of environmental changes.

Predicted highly expressed genes in Nocardia farcinica and the implication for its primary metabolism and nocardial virulence.

AbstractNocardia farcinica is a Gram positive, filamentous bacterium, and is considered an opportunistic pathogen. In this study, the highly expressed genes in N. farcinica were predicted using the codon adaptation index (CAI) as a numerical estimator of gene expressivity. Using ribosomal protein (RP) genes as references, the top ∼n∼10% of the genes were predicted to be the predicted highly expressed (PHX) genes in N. farcinica using a CAI cutoff of greater than 0.73. Consistent with earlier analysis of Streptomyces genomes, most of the PHX genes in N. farcinica were involved in various ‘house-keeping’ functions important for cell growth. However, 15 genes putatively involved in nocardial virulence were predicted as PHX genes in N. farcinica, which included genes encoding four Mce proteins, cyclopropane fatty acid synthase which is involved in the modification of cell wall which may be important for nocardia virulence, polyketide synthase PKS13 for mycolic acid synthesis and a non-ribosomal peptide synthetase involved in biosynthesis of a mycobactin-related siderophore. In addition, multiple genes involved in defense against reactive oxygen species (ROS) produced by the phagocyte were predicted with high expressivity, which included alkylhydroperoxide reductase (ahpC), catalase (katG), superoxide dismutase (sodF), thioredoxin, thioredoxin reductase, glutathione peroxidase, and peptide methionine sulfoxide reductase, suggesting that combating against ROS is essential for survival of N. farcinica in host cells. The study also showed that the distribution of PHX genes in the N. farcinica circular chromosome was uneven, with more PHX genes located in the regions close to replication initiation site. The results provided the first estimates of global gene expression patterns in N.␣farcinica, which will be useful in guiding experimental design for further investigations.

DNA Microarray Analysis of Anaerobic Methanosarcina Barkeri Reveals Responses to Heat Shock and Air Exposure

Methanosarcina barkeri is a methanogenic archaeon that can digest cellulose and other polysaccharides to produce methane. It can only grow under strictly anoxic conditions, but which can survive air exposure. To obtain further knowledge of cellular changes occurring in M. barkeri in response to air exposure and other environmental stresses, we constructed the first oligonucleotide microarray for M. barkeri and used it to investigate the global transcriptomic responses of M. barkeri to air exposure and heat shock at 45°C for 1xa0h. The results showed that various house-keeping genes, such as genes involved in DNA replication recombination and repair, energy production and conversion, and protein turnover were regulated by environmental stimuli. In response to air exposure, up-regulation of a large number of transposase encoding genes was observed. However, no differential expression of genes encoding superoxide dismutase, catalase, nonspecific peroxidases or thioredoxin reductase was observed in response to air exposure, implying that no significant level of reactive oxygen species has been formed under air exposure. In response to heat shock, both Hsp70 (DnaK-DnaJ-GrpE chaperone system) the Hsp60 (GroEL) systems were up-regulated, suggesting that they may play an important role in protein biogenesis in M. barkeri during heat stress.

Microbial dynamics in upflow anaerobic sludge blanket (UASB) bioreactor granules in response to short-term changes in substrate feed

The upflow anaerobic sludge blanket (UASB) reactor is a microcosm for the methanogenic degradation of organic matter in anaerobic environments, and depends on the auto-formation of dense 3D biofilms of 1-3 mm in diameter, referred to as granular sludge (biogranules). Past research has shown that UASB and other methanogenic reactors are extremely stable functionally, but the underlying basis of the functional stability is not well understood. In this study, microbial dynamics in the communities residing in UASB biogranules were analysed to determine responses to short-term perturbations (change in reactor feed). The reactor was fed with simulated brewery wastewater (SBWW) for 1.5 months (phase 1), acetate/sulfate for 2 months (phase 2), acetate alone for 3 months (phase 3) and then a return to SBWW for 2 months (phase 4). Analysis of 16S rRNA, methanogen-associated mcrA and sulfate reducer-associated dsrAB gene-based-clone libraries showed a relatively simple community composed mainly of the methanogenic archaea (Methanobacterium and Methanosaeta), members of the green non-sulfur (Chloroflexi) group of bacteria and Syntrophobacter, Spirochaeta, Acidobacteria and Cytophaga-related bacterial sequences. The mcrA clone libraries were dominated throughout by Methanobacterium- and Methanospirillum-related sequences. Although the reactor performance remained relatively stable throughout the experiment, community diversity levels generally decreased for all libraries in response to a change from SBWW to acetate alone feed. There was a large transitory increase noted in 16S diversity at the 2 month sampling on acetate alone, entirely related to an increase in bacterial diversity. Upon return to SBWW conditions in phase 4, all diversity measures returned to near phase 1 levels. Our results demonstrated that microbial communities, even highly structured ones such as in UASB biogranules, are very capable of responding to rapid and major changes in their environment.

Comparative transcriptome analysis of Desulfovibrio vulgaris grown in planktonic culture and mature biofilm on a steel surface

Biofilm build-up of sulphate-reducing bacteria (SRB) on metal surfaces may lead to severe corrosion of iron. To understand the processes at molecular level, in this study, a whole-genome oligonucleotide microarray was used to examine differential expression patterns between planktonic populations and mature biofilm of Desulfovibrio vulgaris on a steel surface. Statistical analysis revealed that 472 genes were differentially expressed (1.5-fold or more with a q value less than 0.025) by comparing the biofilm cells with the planktonic cells. Among the differentially expressed genes were several that corresponded to genes identified in many aerobic bacterial biofilms (i.e., Pseudomonas species and Escherichia coli) such as genes encoding flagellin, a flagellar motor switch protein, chemotaxis proteins involved in cell motility, as well as genes involved in exopolysaccharide biosynthesis. In addition, the biofilm-bound cells of D. vulgaris exhibited decreased transcription of genes involved in protein synthesis, energy metabolism and sulfate reduction, as well as genes involved in general stress responses. These findings were all consistent with early suggestion that the average physiology of the biofilm cells were similar to cells reduced in growth. Most notably, up-regulation of large number of outer membrane proteins was observed in the D. vulgaris biofilm. Although their function is still unknown, the higher expression of these genes in the biofilm could implicate important roles in the formation and maintenance of multi-cellular consortium on a steel surface. The study provided insights into the metabolic networks associated with the formation and maintenance of a D. vulgaris biofilm on a steel surface.

Two-Component Signal Transduction Systems of Desulfovibrio Vulgaris: Structural and Phylogenetic Analysis and Deduction of Putative Cognate Pairs

A large number of two-component signal transduction systems (TCSTS) including 59 putative sensory histidine kinases (HK) and 55 response regulators (RR) were identified from the Desulfovibrio vulgaris genome. In this study, the structural and phylogenetic analyses of all putative TCSTSs in D. vulgaris were performed. The results showed that D. vulgaris contained 21 hybrid-type HKs, implying that multiple-step phosphorelay may be a common signal transduction mechanism in D. vulgaris. Despite the low sequence similarity that restricted the resolution of the phylogenetic analyses, most TCSTS components of D. vulgaris were found clustered into several subfamilies previously recognized in Escherichia coli and Bacillus subtilis. An attempt was made in this investigation to identify the possible cognate HK-RR pairs not linked on the chromosome in D. vulgaris based on similar expression patterns in response to various environmental and genetic changes. Expression levels of all HK and RR genes were measured using whole-genome microarrays. Five groups of HK-RR genes not linked on the chromosome were identified as possible cognate pairs in D. vulgaris. The results provided a preliminary list of possible cognate HK-RR pairs and constitute a basis for further exploration of interaction and physiological function of TCSTSs in D. vulgaris.