Ensemble modeling for strain development of l-lysine-producing Escherichia coli

One of the main strategies to improve the production of relevant metabolites has been the manipulation of single or multiple key genes in the metabolic pathways. This kind of strategy requires several rounds of experiments to identify enzymes that impact either yield or productivity. The use of mathematical tools to facilitate this process is desirable. In this work, we apply the Ensemble Modeling (EM) framework, which uses phenotypic data (effects of enzyme overexpression or genetic knockouts on the steady-state production rate) to screen for potential models capable of describe existing data and thus gaining insight to improve strains for l-lysine production. Described herein is a strategy to generate a set of kinetic models that describe a set of enzyme overexpression phenotypes previously determined in an Escherichia coli strain that produces increased levels of l-lysine in an industrial laboratory. This final ensemble of models captures the kinetic characteristics of the cell through screening of phenotypes after sequential overexpression of enzymes. Furthermore, these models demonstrate some predictive capability, as starting from the reference producing strain (overexpressing desensitized dihydrodipicolinate synthetase (dapA*)) this set of models is able to predict that the desensitization of aspartate kinase (lysC*) is the next rate-controlling step in the l-lysine pathway. Moreover, this set of models allows for the generation of further targets for testing, for example, phosphoenolpyruvate (Ppc), aspartate aminotransferase (AspC), and glutamate dehydrogenase (GdhA). This work demonstrates the usefulness, applicability, and scope that the Ensemble Modeling framework offers to build production strains.     

Optimization-driven identification of genetic perturbations accelerates the convergence of model parameters in ensemble modeling of metabolic networks

The ensemble modeling (EM) approach has shown promise in capturing kinetic and regulatory effects in the modeling of metabolic networks. Efficacy of the EM procedure relies on the identification of model parameterizations that adequately describe all observed metabolic phenotypes upon perturbation. In this study, we propose an optimization-based algorithm for the systematic identification of genetic/enzyme perturbations to maximally reduce the number of models retained in the ensemble after each round of model screening. The key premise here is to design perturbations that will maximally scatter the predicted steady-state fluxes over the ensemble parameterizations. We demonstrate the applicability of this procedure for an Escherichia coli metabolic model of central metabolism by successively identifying single, double, and triple enzyme perturbations that cause the maximum degree of flux separation between models in the ensemble. Results revealed that optimal perturbations are not always located close to reaction(s) whose fluxes are measured, especially when multiple perturbations are considered. In addition, there appears to be a maximum number of simultaneous perturbations beyond which no appreciable increase in the divergence of flux predictions is achieved. Overall, this study provides a systematic way of optimally designing genetic perturbations for populating the ensemble of models with relevant model parameterizations.     

Pathway engineering for production of aromatics in Escherichia coli: Confirmation of stoichiometric analysis by independent modulation of AroG, TktA, and Pps activities

The synthesis of 3-deoxy-D-arabino-heptulosonate-7-phosphate (DAHP) is the first commitment of resources toward aromatics production in Escherichia coli. DAHP is produced during a condensation reaction between phosphoenolpyruvate (PEP) and erythrose 4-phosphate (E4P) catalyzed by DAHP synthases (coded by aroF, aroG, and aroH). Stoichiometric analysis has shown a severe PEP limitation in the theoretical yield of DAHP production from glucose due to the phosphotransferase system (PTS) for sugar uptake. This limitation can be relieved by (i) the recycling of pyruvate from PEP using PEP synthase (Pps) or (ii) use of non-PTS sugars such as xylose. Previous studies have shown the usefulness of overexpressing tktA (encoding transketolase), aroG, and pps (PEP synthase) for DAHP production in an aroB strain unable to utilize DAHP further. In the present study we confirm the predictions of the stoichiometric analysis by introducing pps, tktA, and aroG into vectors under independently controlled promoters. In glucose medium, although TktA has some positive effect on the final DAHP concentration, it has no effect on the yield (percent conversion). With Pps overexpression, the DAHP concentration produced from glucose is increased almost twofold and the yield is approaching the theoretical maximum, as predicted by the stoichiometric analysis. However, this Pps effect is observed only in the presence of both increased AroG and TktA. In xylose mimimal medium, the final DAHP concentration and the yield are completely determined by the AroG activity. TktA and Pps play no or insignificant roles, and the yield can reach the theoretical maximum without overexpression of these two enzymes. The results shown here are important for both rational design of metabolic pathways and industrial production of aromatics such as tryptophan, phenylalanine, indigo, quinic acid, and catechol.     

Ensemble modeling of metabolic networks

Complete modeling of metabolic networks is desirable, but it is difficult to accomplish because of the lack of kinetics. As a step toward this goal, we have developed an approach to build an ensemble of dynamic models that reach the same steady state. The models in the ensemble are based on the same mechanistic framework at the elementary reaction level, including known regulations, and span the space of all kinetics allowable by thermodynamics. This ensemble allows for the examination of possible phenotypes of the network upon perturbations, such as changes in enzyme expression levels. The size of the ensemble is reduced by acquiring data for such perturbation phenotypes. If the mechanistic framework is approximately accurate, the ensemble converges to a smaller set of models and becomes more predictive. This approach bypasses the need for detailed characterization of kinetic parameters and arrives at a set of models that describes relevant phenotypes upon enzyme perturbations.     

Dealing with Noisy Absences to Optimize Species Distribution Models: An Iterative Ensemble Modelling Approach

Species distribution models (SDMs) are widespread in ecology and conservation biology, but their accuracy can be lowered by non-environmental (noisy) absences that are common in species occurrence data. Here we propose an iterative ensemble modelling (IEM) method to deal with noisy absences and hence improve the predictive reliability of ensemble modelling of species distributions. In the IEM approach, outputs of a classical ensemble model (EM) were used to update the raw occurrence data. The revised data was then used as input for a new EM run. This process was iterated until the predictions stabilized. The outputs of the iterative method were compared to those of the classical EM using virtual species. The IEM process tended to converge rapidly. It increased the consensus between predictions provided by the different methods as well as between those provided by different learning data sets. Comparing IEM and EM showed that for high levels of non-environmental absences, iterations significantly increased prediction reliability measured by the Kappa and TSS indices, as well as the percentage of well-predicted sites. Compared to EM, IEM also reduced biases in estimates of species prevalence. Compared to the classical EM method, IEM improves the reliability of species predictions. It particularly deals with noisy absences that are replaced in the data matrices by simulated presences during the iterative modelling process. IEM thus constitutes a promising way to increase the accuracy of EM predictions of difficult-to-detect species, as well as of species that are not in equilibrium with their environment.     

Stimulation of glucose catabolism in Escherichia coli by a potential futile cycle.

Fifteen-fold overexpression of phosphoenolpyruvate synthase (Pps) (EC 2.7.9.2) in Escherichia coli stimulated oxygen consumption in glucose minimal medium. A further increase in Pps overexpression to 30-fold stimulated glucose consumption by approximately 2-fold and resulted in an increased excretion of pyruvate and acetate. Insertion of two codons at the PvuII site in the pps gene abolished the enzymatic activity and eliminated the above-described effects. Both the active and the inactive proteins were detected at the predicted molecular weight by polyacrylamide gel electrophoresis. Therefore, the observed physiological changes were due to the activity of Pps. The higher specific rates of consumption of oxygen and glucose indicate a potential futile cycle between phosphoenolpyruvate (PEP) and pyruvate. A model for the stimulation of glucose uptake is presented; it involves an increased PEP/pyruvate ratio caused by the overexpressed Pps activity, leading to a stimulation of the PEP:sugar phosphotransferase system.     

Control of gluconeogenic growth by pps and pck in Escherichia coli.

It is well-known that Escherichia coli grows more slowly on gluconeogenic carbon sources than on glucose. This phenomenon has been attributed to either energy or monomer limitation. To investigate this problem further, we varied the expression levels of pck, encoding phosphoenolpyruvate carboxykinase (Pck), and pps, encoding phosphoenolpyruvate synthase (Pps). We found that the growth rates of E. coli in minimal medium supplemented with succinate and with pyruvate are limited by the levels of Pck and Pps, respectively. Optimal overexpression of pck or pps increases the unrestricted growth rates on succinate and on pyruvate, respectively, to the same level attained by the wild-type growth rate on glycerol. Since Pps is needed to supply precursors for biosyntheses, we conclude that E. coli growing on pyruvate is limited by monomer supply. However, because pck is required both for biosyntheses and catabolism for cells growing on succinate, it is possible that growth on succinate is limited by both monomer and energy supplies. The growth yield with respect to oxygen remains approximately constant, even though the overproduction of these enzymes enhances gluconeogenic growth. It appears that the constant yield for oxygen is characteristic of efficient growth on a particular substrate and that the yield is already optimal for wild-type strains. Further increases in either Pck or Pps above the optimal levels become growth inhibitory, and the growth yield for oxygen is reduced, indicating less efficient growth.     

ENU-induced allele of brachyury (Tkt1) exhibits a developmental lethal phenotype similar to the original brachyury (T) mutation

New alleles of brachyury (Tkt1, Tkt4) were induced in the mouse complete tw5 haplotype by ethylnitrosourea (ENU). Like the original brachyury (T) mutation, the new alleles cause a short-tailed phenotype in heterozygotes, and interact with the t complex tail interaction factor (tct) in trans to cause phenotypically tailless mice. Because ENU is mainly a point mutagen, it is important to determine that the new alleles are homozygous embryonic lethal mutations like the original T allele, and to characterize their embryonic lethal phenotype. Moreover, the Tkt1 mutation maps to an inverted position relative to quaking (qk) in t haplotypes as compared with its position on normal chromosome 17. The Tkt1 allele was separated from the resident tw5 lethal gene, tclw5, by recombination, allowing embryology studies to be performed. Embryological analyses show that the Tkt1 allele is nearly identical to the classic T allele. At 9 and 10 days of development, homozygous Tkt1/Tkt1 embryos are grossly abnormal with properties including 1) irregular, disorganized somite pairs, 2) a shortened posterior end of the embryo, 3) an irregular neural tube, and 4) an abnormal notochord. In addition, 10 day-old abnormal embryos have anterior limb buds that point dorsally rather than ventrally, and are smaller than normal littermates. We conclude that the Tkt1 mutation is a valuable allele for both mapping and molecular characterization of the brachyury locus.     

A simple structured model describing the growth of Streptomyces lividans

The growth of Streptomyces lividans in defined media was modeled using a simple structured growth model. Conventional unstructured models like Monod kinetics, substrate inhibition kinetics, and the logistic equation were also used in an attempt to fit the data, but the results were all unsatisfactory. The main reason for failure in applying simple unstructured models is that they cannot describe the long lag phases sometimes observed during growth of S. lividans. The simple structured growth model was derived along similar principles to cybernetic growth models. This model quite accurately describes the growth of S. lividans. It assumes that the rate of assimilation of a substrate depends on the concentration of a specific key enzyme. This key enzyme is only produced in the presence of the substrate, and it is broken down at a steady rate. An enzyme synthesis allocation variable, w, similar to the cybernetic variable, u, described in cybernetic growth models, is proposed to control enzyme synthesis. Until the key enzyme concentration approaches its maximum level, very little substrate is consumed. And consequently, the lag phase is sustained.     

Sensing single base incorporation with nanopore micromanipulation.

Using single-molecule techniques, scientists can routinely investigate the action of a single enzyme. The goal of such studies is usually to gain accurate information unachievable in ensemble assays, such as the maximum instantaneous rate of reaction, the existence of pauses or backward steps, etc. In the article discussed here, the authors have increased their experimental sensitivity so that they can detect a single enzymatic cycle. This improvement makes possible the use of a polymerase enzyme to sequence a single DNA molecule.     

Optimized Two-Stage Ensemble Model for Mammography Mass Recognition

ObjectivesMammography mass recognition is considered as a very challenge pattern recognition problem due to the high similarity between normal and abnormal masses. Therefore, the main objective of this study is to develop an efficient and optimized two-stage recognition model to tackle this recognition task.Material and methodsBasically, the developed recognition model combines an ensemble of linear Support Vector Machine (SVM) classifiers with a Reinforcement Learning-based Memetic Particle Swarm Optimizer (RLMPSO) as RLMPSO-SVM recognition model. RLMPSO is used to construct a two-stage of an ensemble of linear SVM classifiers by performing simultaneous SVM parameters tuning, features selection, and training instances selection. The first stage of RLMPSO-SVM recognition model is responsible about recognizing the input ROI mammography masses as normal or abnormal mass pattern. Meanwhile, the second stage of RLMPSO-SVM model used to perform further recognition for abnormal ROIs as malignant or benign masses. In order to evaluate the effectiveness of RLMPSO-SVM, a total of 1187 normal ROIs, 111 malignant ROIs, and 135 benign ROIs were randomly selected from DDSM database images.ResultsReported results indicated that RLMPSO-SVM model was able to achieve performances of 97.57% sensitivity rate with 97.86% specificity rate for normal vs. abnormal recognition cases. For malignant vs. benign recognition performance it was reported of 97.81% sensitivity rate with 96.92% specificity rate.ConclusionReported results indicated that RLMPSO-SVM recognition model is an effective tool that could assist the radiologist during the diagnosis of the presented abnormalities in mammography images. The outcomes indicated that RLMPSO-SVM significantly outperformed various SVM-based models as well as other variants of computational intelligence models including multi-layer perceptron, naive Bayes classifier, and k-nearest neighbor.     

Metabolic rate and reactive oxygen species production in different genotypes of GH-transgenic zebrafish

Growth hormone overexpression increases growth and consequently increases the metabolic rate in fishes. Therefore, the objective of this study was to evaluate the effects of growth hormone overexpression in zebrafish Danio rerio in terms of growth, oxygen consumption, reactive oxygen species production, lipid hydroperoxide content, antioxidant enzyme activity and glutamate-cysteine ligase catalytic subunit gene expression. The employed models were wild type and transgenic (hemizygous and homozygous) zebrafish expressing the Odonthestes argentinensis growth hormone gene directed by the Cyprinus carpio beta-actin promoter. Higher growth parameters were observed in the hemizygous group. The homozygous group possessed higher oxygen consumption and reactive oxygen species production. Growth hormone transgenesis causes a decrease in glutamate-cysteine ligase catalytic subunit expression, an enzyme responsible for glutathione synthesis. Although the lipid hydroperoxide content was similar between groups, we demonstrate that growth hormone overexpression has the potential to generate oxidative stress in fishes.     

An ensemble model for predicting Rhopalosiphum padi abundance

A novel modeling method is proposed to predict the abundance of the main vector of barley yellow dwarf virus in autumn sown cereal crops. An ensemble model based on artificial neural networks (ANN) was developed to predict the number of Rhopalosiphum padi (L.) (Homoptera: Aphididae) caught in traps during the autumn flight period at Lincoln, Canterbury, New Zealand, over the period 1982–2003. Artificial neural networks were trained using historical weather data and aphid data collected from a suction trap. Model results were compared with those obtained using multiple regression (MR) models using the same independent variables. Both ANN and MR models were validated by leave‐one‐out validation, in other words, by sequentially jackknifing each year out of the data set, fitting a model to the remaining data, then using that model to predict the number of aphids for each jackknifed year. A linear ensemble of ANN models further improved the predictions and represented the trends in the number of aphids over the 22‐year period very well. The r² between the predicted and observed numbers of aphids for the ANN models changed from 0.68 to 0.83 using the linear ensemble model, but the ensemble approach did not change the prediction for the MR models. The absolute mean error (ABSME) of prediction was much lower for the ANN ensemble predictions compared to that for the MR models. The ABMSE for the ANN models dropped from 109 to 86 aphids compared to an ABMSE reduction from 245 to 220 aphids for the MR models. We discuss the potential for ensemble models for predicting insect abundance when long‐term historical data are available.     

The iterative ensemble modelling approach increases the accuracy of fish distribution models

Methodological absences, i.e. when a species is not detected although it is actually present, are known to reduce the prediction accuracy of species distribution models (SDMs). To deal with this problem, we assessed whether a new iterative ensemble modelling (IEM) approach better predicts the spatial distribution of a set of 31 freshwater fish species, exhibiting a wide range of prevalence and methodological absences. Model efficiency was compared using one threshold‐independent (AUC) and three threshold‐dependent indicators of model predictive performance: the percentage of misclassified sites; the Kappa index; and the True Skill Statistic. We then reconstructed species assemblages from individual species predictions and compared observed assemblages to those predicted using EM and IEM using the Jaccard index. Compared to an EM approach, IEM improved model predictive performance for most difficult‐to‐detect species. The iterative approach outperformed EM at modelling the distribution of difficult‐to‐detect species, provided that presence data are representative of the niche of the species. At the assemblage level, the discrepancy between observed and IEM predicted assemblages was significantly lower than that between observed and EM predicted assemblages, showing that IEM can be used to predict the distribution of entire species assemblages. The IEM approach provides a way to consider difficult‐to‐detect species in species distribution models.     

The class II phosphoinositide 3-kinase PI3K-C2beta regulates cell migration by a PtdIns3P dependent mechanism

The biological and pathophysiological significance of class II phosphoinositide 3-kinase enzyme expression currently remains unclear. Using an in vitro scrape wound assay and time-lapse video microscopy, we demonstrate that cell motility is increased in cultures expressing recombinant PI3K-C2beta enzyme. In addition, overexpression of PI3K-C2beta transiently decreased cell adhesion, stimulated the formation of cytoplasmic processes, and decreased the rate of cell proliferation. Consistent with these observations, expression of PI3K-C2beta also decreased expression of alpha4 beta1 integrin subunits. Using asynchronous cultures, we show that endogenous PI3K-C2beta is present in lamellipodia of motile cells. When cells expressing recombinant PI3K-C2beta were plated onto fibronectin, cortical actin staining increased markedly and actin rich lamellipodia and filopodia became evident. Overexpression of a 2xFYVE(Hrs) domain fusion protein abolished this response demonstrating that the effect of PI3K-C2beta on the reorganization of actin filaments is dependent upon PtdIns3P. Finally, overexpression of PI3K-C2beta increased GTP loading of Cdc42. Our data demonstrates for the first time, that PI3K-C2beta plays a regulatory role in cell motility and that the mechanism by which it reorganizes the actin cytoskeleton is dependent upon PtdIns3P production.     

Refining conformational ensembles of flexible proteins against small-angle x-ray scattering data

Intrinsically disordered proteins and flexible regions in multidomain proteins display substantial conformational heterogeneity. Characterizing the conformational ensembles of these proteins in solution typically requires combining one or more biophysical techniques with computational modeling or simulations. Experimental data can either be used to assess the accuracy of a computational model or to refine the computational model to get a better agreement with the experimental data. In both cases, one generally needs a so-called forward model (i.e., an algorithm to calculate experimental observables from individual conformations or ensembles). In many cases, this involves one or more parameters that need to be set, and it is not always trivial to determine the optimal values or to understand the impact on the choice of parameters. For example, in the case of small-angle x-ray scattering (SAXS) experiments, many forward models include parameters that describe the contribution of the hydration layer and displaced solvent to the background-subtracted experimental data. Often, one also needs to fit a scale factor and a constant background for the SAXS data but across the entire ensemble. Here, we present a protocol to dissect the effect of the free parameters on the calculated SAXS intensities and to identify a reliable set of values. We have implemented this procedure in our Bayesian/maximum entropy framework for ensemble refinement and demonstrate the results on four intrinsically disordered proteins and a protein with three domains connected by flexible linkers. Our results show that the resulting ensembles can depend on the parameters used for solvent effects and suggest that these should be chosen carefully. We also find a set of parameters that work robustly across all proteins.     

Differential control of adrenal drug and steroid metabolism in the guinea pig.

Studies were carried out to compare the effects of several physiological variables on adrenal microsomal drug (ethylmorphine demethylation) and steroid (21-hydroxylation) metabolism in guinea pigs. The rate of adrenal ethylmorphine (EM) metabolism increased with maturation in males but not females, resulting in a sex difference (M > F) in adrenal enzyme activity in adult guinea pigs. Twenty-one hydroxylase activity, in contrast, was similar in adrenals from males and females. The concentration of adrenal microsomal cytochrome P-450 was unaffected by age or sex. ACTH administration decreased adrenal EM demethylase activity but did not affect 21-hydroxylation. Testosterone, when given to female guinea pigs, increased the rate of EM metabolism and decreased 21-hydroxylase activity. Various compounds known to interact with adrenal microsomal cytochrome P-450 had divergent effects on EM metabolism and 21-hydroxylation $\text{in}$$\text{vitro}$. Prostaglandins E₁ and F_2α, spironolactone, and canrenone inhibited EM demethylation but not 21-hydroxylation. Simple aromatic hydrocarbons (benzene, toluene), in contrast, inhibited 21-hydroxylation but did not affect EM metabolism. The results indicate that adrenal drug and steroid metabolism are independently regulated and that different terminal oxidases (cytochrome P-450) are probably involved in adrenal 21-hydroxylation and EM demethylation.     

The pimB gene of Mycobacterium tuberculosis encodes a mannosyltransferase involved in lipoarabinomannan biosynthesis.

The biosynthesis of lipoarabinomannan (LAM), a key mycobacterial lipoglycan that has been implicated in numerous immunoregulatory functions, was examined utilizing D-mannosamine (ManN) as a tool to identify mannosyltransferase genes involved in LAM synthesis. Cell-free reactions utilizing cellular membranes of mycobacteria as the enzyme source indicated that ManN inhibited the synthesis of phosphatidylinositol mannosides, early precursors to LAM. A selection strategy was devised to screen a Mycobacterium tuberculosis genomic library in Mycobacterium smegmatis for clones conferring conditional resistance to ManN, with the rationale that overexpression of the gene(s) encoding a target of ManN would impart a ManN-resistant phenotype under these conditions. This strategy led to the identification of pimB, whose deduced amino acid sequence shows similarity to mannosyltransferases and other glycosyltransferases. Partially purified recombinant PimB protein from Escherichia coli or membranes from M. smegmatis overexpressing the pimB gene were used in cell-free assays to show that PimB catalyzes the formation of triacylphosphatidylinositol dimannoside from GDP-mannose and triacylphosphatidylinositol monomannoside.     

Combinatorial pathway analysis for improved L-tyrosine production in Escherichia coli: identification of enzymatic bottlenecks by systematic gene overexpression

Combinatorial overexpression of aromatic amino acid biosynthesis (AAAB) genes in the L-tyrosine producing Escherichia coli strains T1 and T2 was employed to search for AAAB reactions limiting L-tyrosine production. All AAAB genes except aroG and tyrA, which were substituted by their feedback resistant derivatives in the host strains, were cloned and overexpressed. A total of 72 different strains overexpressing various AAAB gene combinations were generated and from those strains with improved phenotype, enzymatic bottlenecks of the AAAB pathway could be inferred. The two major gene overexpression targets for increased L-tyrosine production in E. coli were ydiB and aroK, coding for a shikimate dehydrogenase and a shikimate kinase, respectively, and the combination of ydiB and aroK for overexpression resulted in the best L-tyrosine producing strains in this study, yielding 45% for strain T1 and 26% for strain T2, respectively, higher L-tyrosine titers. Interestingly, overexpression studies with combinations of more than one gene revealed that new gene targets could be identified when overexpessed together with other genes but not alone as single gene overexpression. For example, tyrB encoding the last enzyme of the AAAB pathway, an aromatic amino acid transaminase, improved L-tyrosine production significantly when co-overexpressed together with ydiB or aroK, but not when overexpressed alone. It is also noteworthy that E. coli T1, which generally yielded less L-tyrosine, was amenable to greater improvements than strain T2, i.e. E. coli T1 exhibited generally more space for phenotype improvement.     

Biochemical evaluation of a parsley tyrosine decarboxylase results in a novel 4-hydroxyphenylacetaldehyde synthase enzyme

Plant aromatic amino acid decarboxylases (AAADs) are effectively indistinguishable from plant aromatic acetaldehyde syntheses (AASs) through primary sequence comparison. Spectroscopic analyses of several characterized AASs and AAADs were performed to look for absorbance spectral identifiers. Although this limited survey proved inconclusive, the resulting work enabled the reevaluation of several characterized plant AAS and AAAD enzymes. Upon completion, a previously reported parsley AAAD protein was demonstrated to have AAS activity. Substrate specificity tests demonstrate that this novel AAS enzyme has a unique substrate specificity towards tyrosine (km 0.46mM) and dopa (km 1.40mM). Metabolite analysis established the abundance of tyrosine and absence of dopa in parsley extracts. Such analysis indicates that tyrosine is likely to be the sole physiological substrate. The resulting information suggests that this gene is responsible for the in vivo production of 4-hydroxyphenylacetaldehyde (4-HPAA). This is the first reported case of an AAS enzyme utilizing tyrosine as a primary substrate and the first report of a single enzyme capable of producing 4-HPAA from tyrosine.