Sudhakar Jonnalagadda

Reconstruction and analysis of a genome-scale metabolic model for Scheffersomyces stipitis

BackgroundFermentation of xylose, the major component in hemicellulose, is essential for economic conversion of lignocellulosic biomass to fuels and chemicals. The yeast Scheffersomyces stipitis (formerly known as Pichia stipitis) has the highest known native capacity for xylose fermentation and possesses several genes for lignocellulose bioconversion in its genome. Understanding the metabolism of this yeast at a global scale, by reconstructing the genome scale metabolic model, is essential for manipulating its metabolic capabilities and for successful transfer of its capabilities to other industrial microbes.ResultsWe present a genome-scale metabolic model for Scheffersomyces stipitis, a native xylose utilizing yeast. The model was reconstructed based on genome sequence annotation, detailed experimental investigation and known yeast physiology. Macromolecular composition of Scheffersomyces stipitis biomass was estimated experimentally and its ability to grow on different carbon, nitrogen, sulphur and phosphorus sources was determined by phenotype microarrays. The compartmentalized model, developed based on an iterative procedure, accounted for 814 genes, 1371 reactions, and 971 metabolites. In silico computed growth rates were compared with high-throughput phenotyping data and the model could predict the qualitative outcomes in 74% of substrates investigated. Model simulations were used to identify the biosynthetic requirements for anaerobic growth of Scheffersomyces stipitis on glucose and the results were validated with published literature. The bottlenecks in Scheffersomyces stipitis metabolic network for xylose uptake and nucleotide cofactor recycling were identified by in silico flux variability analysis. The scope of the model in enhancing the mechanistic understanding of microbial metabolism is demonstrated by identifying a mechanism for mitochondrial respiration and oxidative phosphorylation.ConclusionThe genome-scale metabolic model developed for Scheffersomyces stipitis successfully predicted substrate utilization and anaerobic growth requirements. Useful insights were drawn on xylose metabolism, cofactor recycling and mechanism of mitochondrial respiration from model simulations. These insights can be applied for efficient xylose utilization and cofactor recycling in other industrial microorganisms. The developed model forms a basis for rational analysis and design of Scheffersomyces stipitis metabolic network for the production of fuels and chemicals from lignocellulosic biomass.

Analysis of the heat shock response in mouse liver reveals transcriptional dependence on the nuclear receptor peroxisome proliferator-activated receptor α (PPARα)

BackgroundThe nuclear receptor peroxisome proliferator-activated receptor alpha (PPARα) regulates responses to chemical or physical stress in part by altering expression of genes involved in proteome maintenance. Many of these genes are also transcriptionally regulated by heat shock (HS) through activation by HS factor-1 (HSF1). We hypothesized that there are interactions on a genetic level between PPARα and the HS response mediated by HSF1.ResultsWild-type and PPARα-null mice were exposed to HS, the PPARα agonist WY-14,643 (WY), or both; gene and protein expression was examined in the livers of the mice 4 or 24 hrs after HS. Gene expression profiling identified a number of Hsp family members that were altered similarly in both mouse strains. However, most of the targets of HS did not overlap between strains. A subset of genes was shown by microarray and RT-PCR to be regulated by HS in a PPARα-dependent manner. HS also down-regulated a large set of mitochondrial genes specifically in PPARα-null mice that are known targets of PPARγ co-activator-1 (PGC-1) family members. Pretreatment of PPARα-null mice with WY increased expression of PGC-1β and target genes and prevented the down-regulation of the mitochondrial genes by HS. A comparison of HS genes regulated in our dataset with those identified in wild-type and HSF1-null mouse embryonic fibroblasts indicated that although many HS genes are regulated independently of both PPARα and HSF1, a number require both factors for HS responsiveness.ConclusionsThese findings demonstrate that the PPARα genotype has a dramatic effect on the transcriptional targets of HS and support an expanded role for PPARα in the regulation of proteome maintenance genes after exposure to diverse forms of environmental stress including HS.

Metabolic Engineering of a Novel Muconic Acid Biosynthesis Pathway via 4-Hydroxybenzoic Acid in Escherichia coli

ABSTRACT cis,cis-Muconic acid (MA) is a commercially important raw material used in pharmaceuticals, functional resins, and agrochemicals. MA is also a potential platform chemical for the production of adipic acid (AA), terephthalic acid, caprolactam, and 1,6-hexanediol. A strain of Escherichia coli K-12, BW25113, was genetically modified, and a novel nonnative metabolic pathway was introduced for the synthesis of MA from glucose. The proposed pathway converted chorismate from the aromatic amino acid pathway to MA via 4-hydroxybenzoic acid (PHB). Three nonnative genes, pobA, aroY, and catA, coding for 4-hydroxybenzoate hydrolyase, protocatechuate decarboxylase, and catechol 1,2-dioxygenase, respectively, were functionally expressed in E. coli to establish the MA biosynthetic pathway. E. coli native genes ubiC, aroF FBR, aroE, and aroL were overexpressed and the genes ptsH, ptsI, crr, and pykF were deleted from the E. coli genome in order to increase the precursors of the proposed MA pathway. The final engineered E. coli strain produced nearly 170 mg/liter of MA from simple carbon sources in shake flask experiments. The proposed pathway was proved to be functionally active, and the strategy can be used for future metabolic engineering efforts for production of MA from renewable sugars.

Principal components analysis based methodology to identify differentially expressed genes in time-course microarray data

BackgroundTime-course microarray experiments are being increasingly used to characterize dynamic biological processes. In these experiments, the goal is to identify genes differentially expressed in time-course data, measured between different biological conditions. These differentially expressed genes can reveal the changes in biological process due to the change in condition which is essential to understand differences in dynamics.ResultsIn this paper, we propose a novel method for finding differentially expressed genes in time-course data and across biological conditions (say C1 and C2). We model the expression at C1 using Principal Component Analysis and represent the expression profile of each gene as a linear combination of the dominant Principal Components (PCs). Then the expression data from C2 is projected on the developed PCA model and scores are extracted. The difference between the scores is evaluated using a hypothesis test to quantify the significance of differential expression. We evaluate the proposed method to understand differences in two case studies (1) the heat shock response of wild-type and HSF1 knockout mice, and (2) cell-cycle between wild-type and Fkh1/Fkh2 knockout Yeast strains.ConclusionIn both cases, the proposed method identified biologically significant genes.

NIFTI: An evolutionary approach for finding number of clusters in microarray data

BackgroundClustering techniques are routinely used in gene expression data analysis to organize the massive data. Clustering techniques arrange a large number of genes or assays into a few clusters while maximizing the intra-cluster similarity and inter-cluster separation. While clustering of genes facilitates learning the functions of un-characterized genes using their association with known genes, clustering of assays reveals the disease stages and subtypes. Many clustering algorithms require the user to specify the number of clusters a priori. A wrong specification of number of clusters generally leads to either failure to detect novel clusters (disease subtypes) or unnecessary splitting of natural clusters.ResultsWe have developed a novel method to find the number of clusters in gene expression data. Our procedure evaluates different partitions (each with different number of clusters) from the clustering algorithm and finds the partition that best describes the data. In contrast to the existing methods that evaluate the partitions independently, our procedure considers the dynamic rearrangement of cluster members when a new cluster is added. Partition quality is measured based on a new index called Net InFormation Transfer Index (NIFTI) that measures the information change when an additional cluster is introduced. Information content of a partition increases when clusters do not intersect and decreases if they are not clearly separated. A partition with the highest Total Information Content (TIC) is selected as the optimal one. We illustrate our method using four publicly available microarray datasets.ConclusionIn all four case studies, the proposed method correctly identified the number of clusters and performs better than other well known methods. Our method also showed invariance to the clustering techniques.

Transcriptomic Analysis of 3-Hydroxypropanoic Acid Stress in Escherichia coli

The stress response of Escherichia coli to 3-hydroxypropanoic acid (3-HP) was elucidated through global transcriptomic analysis. Around 375 genes showed difference of more than 2-fold in 3-HP-treated samples. Further analysis revealed that the toxicity effect of 3-HP was due to the cation and anion components of this acid and some effects-specific to 3-HP. Genes related to the oxidative stress, DNA protection, and repair were upregulated in treated cells due to the lowered cytoplasmic pH caused by accumulated cations. 3-HP-treated E. coli used the arginine acid tolerance mechanism to increase the cytoplasmic pH. Additionally, the anion effects were manifested as imbalance in the osmotic pressure. Analysis of top ten highly upregulated genes suggests the formation of 3-hydroxypropionaldehyde under 3-HP stress. The transcriptomic analysis shed light on the global genetic reprogramming due to 3-HP stress and suggests strategies for increasing the tolerance of E. coli toward 3-HP.

Graph theory augmented math programming approach to identify minimal reaction sets in metabolic networks

Abstract Bioprocesses are of growing importance as an avenue to produce chemicals. Microorganisms containing only desired catalytic and replication capabilities in their metabolic pathways are expected to offer efficient processes for chemical production. Realizing such minimal cells is the holy grail of metabolic engineering. In this paper, we propose a new method that combines graph-theoretic approaches with mixed-integer liner programming (MILP) to design metabolic networks with minimal reactions. Existing MILP based computational approaches are computationally complex especially for large networks. The proposed graph-theoretic approach offers an efficient divide-and-conquer strategy using the MILP formulation on sub-networks rather than considering the whole network monolithically. In addition to the resulting improvement in computational complexity, the proposed method also aids in identifying the key reactions to be knocked-out in order to achieve the minimal cell. The efficacy of the proposed approach is demonstrated using three case studies from two organisms, Escherichia coli and Saccharomyces cerevisiae .

A PCA-Based approach for gene target selection to improve industrial strains

Abstract The production of recombinant proteins has become indispensable for both research and industrial applications. However, the expression of recombinant protein acts as a stress on host strain, resulting in decrease in the rate of growth and hence the productivity of the protein. To improve yield, it is essential to understand the changes in the physiology and metabolism of the host and reverse them by over- or under-expressing the key genes. In this paper, we propose an approach based on Principal Component Analysis to identify the genes differentially expressed in the host strain compared to wild-type strain. These genes provide the information about the changes in the metabolic events due to recombinant protein production. Our approach also identifies the regulators responsible for these changes and hence by over-expressing or knocking-out these regulators, the behavior of the host can be brought to normal. We illustrate the proposed approach using a case study of recombinant protein production in E coli.

Graph Theory Augmented Recursive MILP Approach for Identifying Multiple Minimal Reaction Sets in Metabolic Networks

Abstract Development of cells with minimal functionality, containing only desired catalytic properties for chemical conversion and replication, are gaining importance since such minimal cells are expected to be the most efficient machinery for production of specific chemicals. In this paper, we propose a graph theory augmented recursive MILP approach to identify multiple minimal reaction sets in metabolic networks that are capable of satisfying predefined objectives (such as growth). The proposed approach uses graph theoretic insights to reduce computational time and a recursive MILP approach to identify multiple minimal reaction sets. Identifying such multiple minimal reaction sets facilitates development of best minimal cell based on other process requirements. The proposed approach is illustrated by identifying multiple minimal reaction sets that can produce predefined biomass in E.coli.

A Graph Theory Augmented Math Programming Approach to Identify Genetic Targets for Strain Improvement

Abstract Improvement of biological strains through targeted modification of metabolism is essential for successful development of bioprocesses. The computational complexity of optimization procedures routinely used for identifying genetic targets limits their application to genome-scale metabolic networks. In this study, we combined graph theoretic approaches with mixed-integer liner programming (MILP) to reduce the search space and thus reducing computational time. Specifically, we used cut-sets (minimal set of reactions that cuts metabolic networks) as additional constraints to reduce the search space. The efficacy of proposed approach is illustrated by identifying minimal reaction set for Saccharomyces Cerevisiae.