Reducing phenotypic instabilities of a microbial population during continuous cultivation based on cell switching dynamics.

Predicting the fate of individual cells among a microbial population (i.e., growth and gene expression) remains a challenge, especially when this population is exposed to very dynamic environmental conditions, such as those encountered during continuous cultivation. Indeed, the dynamic nature of a continuous cultivation process implies the potential diversification of the microbial population resulting in genotypic and phenotypic heterogeneity. The present work focused on the induction of the arabinose operon in Escherichia coli as a model system to study this diversification process in continuous cultivations. As a preliminary step, the green fluorescent protein (GFP) level triggered by an arabinose-inducible ParaBAD promoter was tracked by flow cytometry in chemostat cultivations with glucose-arabinose co-feeding. For a wide range of glucose-arabinose co-feeding concentrations in the chemostats, the simultaneous occurrence of GFP positive and negative subpopulation was observed. In the second set of experiments, continuous cultivation was performed by adding glucose continuously and arabinose based on the capability of individual cells to switch from low GFP to high GFP expression states, performed with a technology setup called segregostat. In the segregostat cultivation mode, on-line flow cytometry analysis was used for adjusting the arabinose/glucose transitions based on the phenotypic switching profiles of the microbial population. This strategy allowed finding an appropriate arabinose pulsing frequency, leading to prolonged maintenance of the induction level with a limited increase in the phenotypic diversity for more than 60 generations. The results suggest that the steady forcing of individual cells into a given phenotypic trajectory may not be the best strategy for controlling cell populations. Instead, allowing individual cells to switch periodically around a predefined threshold seems to be a more robust strategy leading to oscillations, but within a predictable cell population behavior range.

Engineering AraC to make it responsive to light instead of arabinose.

The L-arabinose-responsive AraC and its cognate PBAD promoter underlie one of the most often used chemically inducible prokaryotic gene expression systems in microbiology and synthetic biology. Here, we change the sensing capability of AraC from L-arabinose to blue light, making its dimerization and the resulting PBAD activation light-inducible. We engineer an entire family of blue light-inducible AraC dimers in Escherichia coli (BLADE) to control gene expression in space and time. We show that BLADE can be used with pre-existing L-arabinose-responsive plasmids and strains, enabling optogenetic experiments without the need to clone. Furthermore, we apply BLADE to control, with light, the catabolism of L-arabinose, thus externally steering bacterial growth with a simple transformation step. Our work establishes BLADE as a highly practical and effective optogenetic tool with plug-and-play functionality-features that we hope will accelerate the broader adoption of optogenetics and the realization of its vast potential in microbiology, synthetic biology and biotechnology.

Conversion of L-arabinose to L-ribose by genetically engineered Candida tropicalis.

L-Ribose, a starting material for the synthesis of L-nucleoside, has attracted lots of attention since L-nucleoside is responsible for the antiviral activities of the racemic mixtures of nucleoside enantiomers. In this study, the L-ribulose-producing Candida tropicalis strain was engineered for the conversion of L-arabinose to L-ribose. For the construction of a uracil auxotroph, the URA3 gene was excised by homologous recombination. The expression cassette of codon-optimized L-ribose isomerase gene from Acinetobacter calcoaceticus DL-28 under the control of the GAPDH promoter was integrated to the uracil auxotroph. The resulting strain, K1 CoSTP2 LsaAraA AcLRI, was cultivated with the glucose/L-arabinose mixture. At 45.5 h of fermentation, 6.0 g/L of L-ribose and 3.2 g/L of L-ribulose were produced from 30 g/L of L-arabinose. The proportion between L-ribose and L-ribulose was approximately 2:1 and the conversion yield of L-arabinose to L-ribose was about 20% (w/w). The L-ribose-producing yeast strain was successfully constructed for the first time and could convert L-arabinose to L-ribose in one-pot fermentation using the mixture of glucose and L-arabinose.

Biosynthesis of polyhydroxyalkanoates from vegetable oil under the co-expression of fadE and phaJ genes in Cupriavidus necator.

The acyl-CoA dehydrogenase (FadE) and (R)-specific enoyl-CoA hydratase (PhaJ) are functionally related to the degradation of fatty acids and the synthesis of polyhydroxyalkanoates (PHAs). To verify this, a recombinant Cupriavidus necator H16 harboring the plasmid -pMPJAS03- with fadE from Escherichia coli strain K12 and phaJ1 from Pseudomonas putida strain KT2440 under the arabinose promoter (araC-PBAD) was constructed. The impact of co-expressing fadE and phaJ genes on C. necator H16/pMPJAS03 maintaining the wild-type synthase on short-chain-length/medium-chain-length PHA formation from canola or avocado oil at different arabinose concentrations was investigated. The functional activity of fadEE.c led to obtaining higher biomass and PHA concentrations compared to the cultures without expressing the gene. While high transcriptional levels of phaJ1P.p, at 0.1% of arabinose, aid the wild-type synthase to polymerize larger-side chain monomers, such as 3-Hydroxyoctanoate (3HO) and 3-Hydroxydecanoate (3HD). The presence of even small amounts of 3HO and 3HD in the co-polymers significantly depresses the melting temperature of the polymers, compared to those composed of pure 3-hydroxybutyrate (3HB). Our data presents supporting evidence that the synthesis of larger-side chain monomers by the recombinant strain relies not only upon the affinity of the wild-type synthase but also on the functionality of the intermediate supplying enzymes.

A novel DMAPP-responding genetic circuit sensor for high-throughput screening and evolving isoprene synthase.

High-throughput screening is a popular tool for collating biological data which would otherwise require the use of excessive resources. In this study, an artificial genetic circuit sensor responding to dimethylallyl diphosphate (DMAPP) was constructed based on a modified L-arabinose operon for high-throughput screening and isoprene synthase (ispS) evolution in Escherichia coli (E. coli). As a first step, the DNA sequence of the L-arabinose ligand-binding domain (LBD) was replaced with an ispS gene to enable the AraC operon responding to DMAPP, which is the substrate of the IspS enzyme. Then, an enhanced GFP (eGFP) was also introduced as a reporter for pBAD promoter. The expression level of the reporter was monitored using either of the two tools: flow cytometer (FCM) and microplate reader. Sequentially, we observed that a high DMAPP concentration led to low eGFP fluorescence, and the overexpression of ispS gene, which consumes DMAPP, resulted in a high eGFP expression. These results demonstrated that the artificial genetic circuit sensor responded directly to the intracellular concentration of DMAPP, and the expression of IspS enzyme could be positively correlated to the expression level of eGFP. Finally, we identified two IspS mutants with different activities from an ispS gene library and further validated the screening method.

Combinatorial control of gene expression in Aspergillus niger grown on sugar beet pectin.

Aspergillus niger produces an arsenal of extracellular enzymes that allow synergistic degradation of plant biomass found in its environment. Pectin is a heteropolymer abundantly present in the primary cell wall of plants. The complex structure of pectin requires multiple enzymes to act together. Production of pectinolytic enzymes in A. niger is highly regulated, which allows flexible and efficient capture of nutrients. So far, three transcriptional activators have been linked to regulation of pectin degradation in A. niger. The L-rhamnose-responsive regulator RhaR controls the production of enzymes that degrade rhamnogalacturonan-I. The L-arabinose-responsive regulator AraR controls the production of enzymes that decompose the arabinan and arabinogalactan side chains of rhamnogalacturonan-II. The D-galacturonic acid-responsive regulator GaaR controls the production of enzymes that act on the polygalacturonic acid backbone of pectin. This project aims to better understand how RhaR, AraR and GaaR co-regulate pectin degradation. For that reason, we constructed single, double and triple disruptant strains of these regulators and analyzed their growth phenotype and pectinolytic gene expression in A. niger grown on sugar beet pectin.

Bistability and Asynchrony in a Boolean Model of the L-arabinose Operon in Escherichia coli.

The lactose operon in Escherichia coli was the first known gene regulatory network, and it is frequently used as a prototype for new modeling paradigms. Historically, many of these modeling frameworks use differential equations. More recently, Stigler and Veliz-Cuba proposed a Boolean model that captures the bistability of the system and all of the biological steady states. In this paper, we model the well-known arabinose operon in E. coli with a Boolean network. This has several complex features not found in the lac operon, such as a protein that is both an activator and repressor, a DNA looping mechanism for gene repression, and the lack of inducer exclusion by glucose. For 11 out of 12 choices of initial conditions, we use computational algebra and Sage to verify that the state space contains a single fixed point that correctly matches the biology. The final initial condition, medium levels of arabinose and no glucose, successfully predicts the system's bistability. Finally, we compare the state space under synchronous and asynchronous update and see that the former has several artificial cycles that go away under a general asynchronous update.

Two-component systems required for virulence in Pseudomonas aeruginosa.

Pseudomonas aeruginosa is a versatile opportunistic pathogen capable of infecting a broad range of hosts, in addition to thriving in a broad range of environmental conditions outside of hosts. With this versatility comes the need to tightly regulate its genome to optimise its gene expression and behaviour to the prevailing conditions. Two-component systems (TCSs) comprising sensor kinases and response regulators play a major role in this regulation. This minireview discusses the growing number of TCSs that have been implicated in the virulence of P. aeruginosa, with a special focus on the emerging theme of multikinase networks, which are networks comprising multiple sensor kinases working together, sensing and integrating multiple signals to decide upon the best response. The networks covered in depth regulate processes such as the switch between acute and chronic virulence (GacS network), the Cup fimbriae (Roc network and Rcs/Pvr network), the aminoarabinose modification of lipopolysaccharide (a network involving the PhoQP and PmrBA TCSs), twitching motility and virulence (a network formed from the Chp chemosensory pathway and the FimS/AlgR TCS), and biofilm formation (Wsp chemosensory pathway). In addition, we highlight the important interfaces between these systems and secondary messenger signals such as cAMP and c-di-GMP.

Protein depletion using the arabinose promoter in Xanthomonas citri subsp. citri.

Xanthomonas citri subsp. citri (X. citri) is a plant pathogen and the etiological agent of citrus canker, a severe disease that affects all the commercially important citrus varieties, and has worldwide distribution. Citrus canker cannot be healed, and the best method known to control the spread of X. citri in the orchards is the eradication of symptomatic and asymptomatic plants in the field. However, in the state of São Paulo, Brazil, the main orange producing area in the world, control is evolving to an integrated management system (IMS) in which growers have to use less susceptible plants, windshields to prevent bacterial spread out and sprays of cupric bactericidal formulations. Our group has recently proposed alternative methods to control citrus canker, which are based on the use of chemical compounds able to disrupt vital cellular processes of X. citri. An important step in this approach is the genetic and biochemical characterization of genes/proteins that are the possible targets to be perturbed, a task not always simple when the gene/protein under investigation is essential for the organism. Here, we describe vectors carrying the arabinose promoter that enable controllable protein expression in X. citri. These vectors were used as complementation tools for the clean deletion of parB in X. citri, a widespread and conserved gene involved in the essential process of bacterial chromosome segregation. Overexpression or depletion of ParB led to increased cell size, which is probably a resultant of delayed chromosome segregation with subsequent retard of cell division. However, ParB is not essential in X. citri, and in its absence the bacterium was fully competent to colonize the host citrus and cause disease. The arabinose expression vectors described here are valuable tools for protein expression, and especially, to assist in the deletion of essential genes in X. citri.

Variation in Mutational Robustness between Different Proteins and the Predictability of Fitness Effects.

Random mutations in genes from disparate protein classes may have different distributions of fitness effects (DFEs) depending on different structural, functional, and evolutionary constraints. We measured the fitness effects of 156 single mutations in the genes encoding AraC (transcription factor), AraD (enzyme), and AraE (transporter) used for bacterial growth on l-arabinose. Despite their different molecular functions these genes all had bimodal DFEs with most mutations either being neutral or strongly deleterious, providing a general expectation for the DFE. This contrasts with the unimodal DFEs previously obtained for ribosomal protein genes where most mutations were slightly deleterious. Based on theoretical considerations, we suggest that the 33-fold higher average mutational robustness of ribosomal proteins is due to stronger selection for reduced costs of translational and transcriptional errors. Whereas the large majority of synonymous mutations were deleterious for ribosomal proteins genes, no fitness effects could be detected for the AraCDE genes. Four mutations in AraC and AraE increased fitness, suggesting that slightly advantageous mutations make up a significant fraction of the DFE, but that they often escape detection due to the limited sensitivity of commonly used fitness assays. We show that the fitness effects of amino acid substitutions can be predicted based on evolutionary conservation, but those weakly deleterious mutations are less reliably detected. This suggests that large-effect mutations and the fraction of highly deleterious mutations can be computationally predicted, but that experiments are required to characterize the DFE close to neutrality, where many mutations ultimately fixed in a population will occur.

The Cell Wall Arabinose-Deficient Arabidopsis thaliana Mutant murus5 Encodes a Defective Allele of REVERSIBLY GLYCOSYLATED POLYPEPTIDE2.

Traditional marker-based mapping and next-generation sequencing was used to determine that the Arabidopsis (Arabidopsis thaliana) low cell wall arabinose mutant murus5 (mur5) encodes a defective allele of REVERSIBLY GLYCOSYLATED POLYPEPTIDE2 (RGP2). Marker analysis of 13 F2 confirmed mutant progeny from a recombinant mapping population gave a rough map position on the upper arm of chromosome 5, and deep sequencing of DNA from these 13 lines gave five candidate genes with G→A (C→T) transitions predicted to result in amino acid changes. Of these five, only insertional mutant alleles of RGP2, a gene that encodes a UDP-arabinose mutase that interconverts UDP-arabinopyranose and UDP-arabinofuranose, exhibited the low cell wall arabinose phenotype. The identities of mur5 and two SALK insertional alleles were confirmed by allelism tests and overexpression of wild-type RGP2 complementary DNA placed under the control of the 35S promoter in the three alleles. The mur5 mutation results in the conversion of cysteine-257 to tyrosine-257 within a conserved hydrophobic cluster predicted to be distal to the active site and essential for protein stability and possible heterodimerization with other isoforms of RGP.

Impact of Residual Inducer on Titratable Expression Systems.

Inducible expression systems are widely employed for the titratable control of gene expression, yet molecules inadvertently present in the growth medium or synthesized by the host cells can alter the response profile of some of these systems. Here, we explored the quantitative impact of these residual inducers on the apparent response properties of inducible systems. Using a simple mathematical model, we found that the presence of residual inducer shrinks the apparent dynamic range and causes the apparent Hill coefficient to converge to one. We also found that activating systems were more sensitive than repressing systems to the presence of residual inducer and the response parameters were most heavily dependent on the original Hill coefficient. Experimental interrogation of common titratable systems based on an L-arabinose inducible promoter or a thiamine pyrophosphate-repressing riboswitch in Escherichia coli confirmed the predicted trends. We finally found that residual inducer had a distinct effect on "all-or-none" systems, which exhibited increased sensitivity to the added inducer until becoming fully induced. Our findings indicate that residual inducer or repressor alters the quantitative response properties of titratable systems, impacting their utility for scientific discovery and pathway engineering.

Escherichia coli as host for membrane protein structure determination: a global analysis.

The structural biology of membrane proteins (MP) is hampered by the difficulty in producing and purifying them. A comprehensive analysis of protein databases revealed that 213 unique membrane protein structures have been obtained after production of the target protein in E. coli. The primary expression system used was the one based on the T7 RNA polymerase, followed by the arabinose and T5 promoter based expression systems. The C41λ(DE3) and C43λ(DE3) bacterial mutant hosts have contributed to 28% of non E. coli membrane protein structures. A large scale analysis of expression protocols demonstrated a preference for a combination of bacterial host-vector together with a bimodal distribution of induction temperature and of inducer concentration. Altogether our analysis provides a set of rules for the optimal use of bacterial expression systems in membrane protein production.

Comparative analysis of the secretion capability of early and late flagellar type III secretion substrates.

A remarkable feature of the flagellar-specific type III secretion system (T3SS) is the selective recognition of a few substrate proteins among the many thousand cytoplasmic proteins. Secretion substrates are divided into two specificity classes: early substrates secreted for hook-basal body (HBB) construction and late substrates secreted after HBB completion. Secretion was reported to require a disordered N-terminal secretion signal, mRNA secretion signals within the 5'-untranslated region (5'-UTR) and for late substrates, piloting proteins known as the T3S chaperones. Here, we utilized translational ß-lactamase fusions to probe the secretion efficacy of the N-terminal secretion signal of fourteen secreted flagellar substrates in Salmonella enterica. We observed a surprising variety in secretion capability between flagellar proteins of the same secretory class. The peptide secretion signals of the early-type substrates FlgD, FlgF, FlgE and the late-type substrate FlgL were analysed in detail. Analysing the role of the 5'-UTR in secretion of flgB and flgE revealed that the native 5'-UTR substantially enhanced protein translation and secretion. Based on our data, we propose a multicomponent signal that drives secretion via the flagellar T3SS. Both mRNA and peptide signals are recognized by the export apparatus and together with substrate-specific chaperones allowing for targeted secretion of flagellar substrates.

Extra- and intracellular lactose catabolism in Penicillium chrysogenum: phylogenetic and expression analysis of the putative permease and hydrolase genes.

Penicillium chrysogenum is used as an industrial producer of penicillin. We investigated its catabolism of lactose, an abundant component of whey used in penicillin fermentation, comparing the type strain NRRL 1951 with the high producing strain AS-P-78. Both strains grew similarly on lactose as the sole carbon source under batch conditions, exhibiting almost identical time profiles of sugar depletion. In silico analysis of the genome sequences revealed that P. chrysogenum features at least five putative ß-galactosidase (bGal)-encoding genes at the annotated loci Pc22g14540, Pc12g11750, Pc16g12750, Pc14g01510 and Pc06g00600. The first two proteins appear to be orthologs of two Aspergillus nidulans family 2 intracellular glycosyl hydrolases expressed on lactose. The latter three P. chrysogenum proteins appear to be distinct paralogs of the extracellular bGal from A. niger, LacA, a family 35 glycosyl hydrolase. The P. chrysogenum genome also specifies two putative lactose transporter genes at the annotated loci Pc16g06850 and Pc13g08630. These are orthologs of paralogs of the gene encoding the high-affinity lactose permease (lacpA) in A. nidulans for which P. chrysogenum appears to lack the ortholog. Transcript analysis of Pc22g14540 showed that it was expressed exclusively on lactose, whereas Pc12g11750 was weakly expressed on all carbon sources tested, including D-glucose. Pc16g12750 was co-expressed with the two putative intracellular bGal genes on lactose and also responded on L-arabinose. The Pc13g08630 transcript was formed exclusively on lactose. The data strongly suggest that P. chrysogenum exhibits a dual assimilation strategy for lactose, simultaneously employing extracellular and intracellular hydrolysis, without any correlation to the penicillin-producing potential of the studied strains.

Improvement of L-arabinose fermentation by modifying the metabolic pathway and transport in Saccharomyces cerevisiae.

The L-arabinose utilization pathway was established in Saccharomyces cerevisiae, by expressing the codon-optimized araA, araB, and araD genes of Lactobacillus plantarum. After overexpressing the TAL1, TKL1, RPE1, RKI1, and GAL2 genes and adaptive evolution, the L-arabinose utilization of the recombinant strain became efficient. The resulting strain displayed a maximum specific growth rate of 0.075 h(-1), a maximum specific L-arabinose consumption rate of 0.61 g h(-1) g(-1) dry cell weight, and a promising ethanol yield of 0.43 g g(-1) from L-arabinose fermentation.

A ß-glucuronosyltransferase from Arabidopsis thaliana involved in biosynthesis of type II arabinogalactan has a role in cell elongation during seedling growth.

We have characterized a ß-glucuronosyltransferase (AtGlcAT14A) from Arabidopsis thaliana that is involved in the biosynthesis of type II arabinogalactan (AG). This enzyme belongs to the Carbohydrate Active Enzyme database glycosyltransferase family 14 (GT14). The protein was localized to the Golgi apparatus when transiently expressed in Nicotiana benthamiana. The soluble catalytic domain expressed in Pichia pastoris transferred glucuronic acid (GlcA) to ß-1,6-galactooligosaccharides with degrees of polymerization (DP) ranging from 3-11, and to ß-1,3-galactooligosaccharides of DP5 and 7, indicating that the enzyme is a glucuronosyltransferase that modifies both the ß-1,6- and ß-1,3-galactan present in type II AG. Two allelic T-DNA insertion mutant lines showed 20-35% enhanced cell elongation during seedling growth compared to wild-type. Analyses of AG isolated from the mutants revealed a reduction of GlcA substitution on Gal-ß-1,6-Gal and ß-1,3-Gal, indicating an in vivo role of AtGlcAT14A in synthesis of those structures in type II AG. Moreover, a relative increase in the levels of 3-, 6- and 3,6-linked galactose (Gal) and reduced levels of 3-, 2- and 2,5-linked arabinose (Ara) were seen, suggesting that the mutation in AtGlcAT14A results in a relative increase of the longer and branched ß-1,3- and ß-1,6-galactans. This increase of galactosylation in the mutants is most likely caused by increased availability of the O6 position of Gal, which is a shared acceptor site for AtGlcAT14A and galactosyltransferases in synthesis of type II AG, and thus addition of GlcA may terminate Gal chain extension. We discuss a role for the glucuronosyltransferase in the biosynthesis of type II AG, with a biological role during seedling growth.

Characterization of a recombinant bifunctional xylosidase/arabinofuranosidase from Phanerochaete chrysosporium.

A bifunctional xylosidase/arabinofuranosidase gene (PcXyl) was cloned from the cDNA library of Phanerochaete chrysosporium and further expressed in Pichia pastoris. Enzymatic assay indicated that P. pastoris produced rPcXyl at a level of 26,141 U l⁻¹. The xylosidase and arabinofuranosidase activities of rPcXyl were maximized, respectively, at pHs of 5.0 and 5.5 and temperatures of 45°C and 50°C. SDS-PAGE revealed a single band of purified rPcXyl of 83 kDa. Cu²âº and Zn²âº completely inhibited the enzyme activity of rPcXyl. The enzyme activity of rPcXyl was increased 151%, 126% and 123%, respectively, in the presence of glucose, xylose and arabinose at concentrations of 5 mM. rPcXyl hydrolyzed xylobiose to xylose and xylotriose to xylose and xylobiose, indicating rPcXyl acts as an exo-type enzyme. Additionally, rPcXyl enhanced xylose release from xylan substrates in synergy with rPcXynC.

A small multidrug resistance-like transporter involved in the arabinosylation of arabinogalactan and lipoarabinomannan in mycobacteria.

The biosynthesis of the major cell envelope glycoconjugates of Mycobacterium tuberculosis is topologically split across the plasma membrane, yet nothing is known of the transporters required for the translocation of lipid-linked sugar donors and oligosaccharide intermediates from the cytoplasmic to the periplasmic side of the membrane in mycobacteria. One of the mechanisms used by prokaryotes to translocate lipid-linked phosphate sugars across the plasma membrane relies on translocases that share resemblance with small multidrug resistance transporters. The presence of an small multidrug resistance-like gene, Rv3789, located immediately upstream from dprE1/dprE2 responsible for the formation of decaprenyl-monophosphoryl-ß-D-arabinose (DPA) in the genome of M. tuberculosis led us to investigate its potential involvement in the formation of the major arabinosylated glycopolymers, lipoarabinomannan (LAM) and arabinogalactan (AG). Disruption of the ortholog of Rv3789 in Mycobacterium smegmatis resulted in a reduction of the arabinose content of both AG and LAM that accompanied the accumulation of DPA in the mutant cells. Interestingly, AG and LAM synthesis was restored in the mutant not only upon expression of Rv3789 but also upon that of the undecaprenyl phosphate aminoarabinose flippase arnE/F genes from Escherichia coli. A bacterial two-hybrid system further indicated that Rv3789 interacts in vivo with the galactosyltransferase that initiates the elongation of the galactan domain of AG. Biochemical and genetic evidence is thus consistent with Rv3789 belonging to an AG biosynthetic complex, where its role is to reorient DPA to the periplasm, allowing this arabinose donor to then be used in the buildup of the arabinan domains of AG and LAM.

State-selective metabolic labeling of cellular proteins.

Transcriptional activity from a specified promoter can provide a useful marker for the physiological state of a cell. Here we introduce a method for selective tagging of proteins made in cells in which specified promoters are active. Tagged proteins can be modified with affinity reagents for enrichment or with fluorescent dyes for visualization. The method allows state-selective analysis of the proteome, whereby proteins synthesized in predetermined physiological states can be identified. The approach is demonstrated by proteome-wide labeling of bacterial proteins upon activation of the P(BAD) promoter and the SoxRS regulon and provides a basis for analysis of more complex systems including spatially heterogeneous microbial cultures and biofilms.