Search

Additional file 1: Figure. S1 Cell growth and description of galR mutations detected in pfk_ALE-2, -3, and -5 strains. (a) Growth rate profiles (OD600) were compared for the four evolved strains isolated after the 50th subculture during ALE. Growth rate profiles of ΔpfkAB and ALE-1 strains are presented as controls. (b, c) DNA sequence of wild-type and mutant galR coding region in ALE-2, -3 and -5 strains were compared. The sequences include the start codon (lower case), emerging site of early stop codon (grey box), duplicated target sequence (underlined), duplication (red), and insertion (blue). Adenine in the start codon of the galR gene is designated as +1. (b) A nonsense mutation was caused by a frame shift due to a 7-bp duplication in the galR gene of the ALE-2, -3 strains, which resulted in truncation of the C-terminal 33 amino acids of the total 344 amino acids. (c) Insertion of 768 bp and duplication of 8 bp in the galR gene of ALE-5 caused insertion of 259 amino acids between the 3rd and 4th amino acids of the wild-type GalR protein. These GalR mutants are expected to be equivalent to the GalR mutant of ALE-1, which has low binding affinity to target sequence in the absence of an appropriate inducer, galactose. Figure. S2 Cell growth and 3-HP, glucose, and acetate concentration during 3-HP production. All strains have a fabI temperature-sensitive mutant (fabIts) and two plasmids pMCR and pACC. Profiles of 3-HP concentration (red circles), cell mass (yellow triangles), consumed glucose concentration (green squares) and acetate (blue diamonds) were compared for E. coli strains: (a) MG1655 (b) MG1655 ∆pfkA (c) MG1655 ∆pgi (d) ΔpfkAB (e) ALE-1 (f) ALE-1 with pGnd and (g) ALE-1 with pGnd-GntR. Cell mass was calculated from the OD600 values. 1 OD600 unit corresponds to 0.33 g CDW/L [1]. Error bars indicate standard deviations of three independent biological replicates. Figure. S3 The comparison of cell growth between pfk_ALE-1 and pfk_ALE-1 Δedd strains. Growth rate profiles (OD600) of ALE-1 Δedd strain was compared to that of ALE-1 strain on glucose as a sole carbon source. Cells were grown in 250 mL with 25 mL of M9 minimal medium containing 0.4% glucose. Error bars indicate standard deviations of three independent biological replicates. Figure .S4 Functional analysis of mutations involved in glucose uptake system in the evolved strains. (a) Target genes affected by the mutated TF (GalR) or ptsG deletion were identified and their role for improved cell growth in ALE-1 were determined. Cells were grown in M9 minimal medium with 0.4% glucose in shaking flasks. IPTG (6.25 µM) was added at the beginning of seeding, subculturing, and culturing to induce each plasmid. Error bars indicate standard deviations of three independent biological replicates. (b) The schematic diagram illustrates the altered route of glucose uptake in ALE-1 via a non-PTS glucose transporter GalP, the expression of which is negatively regulated by GalR. Figure. S5 Comparison of mRNA expression level of galP and talB in ALE-1 with MG1655. The mRNA expression level of galP and talB in MG1655 (gray bar) and ALE-1 (black bar) was analyzed by quantitative RT-PCR. Error bars indicate standard deviations of three independent biological replicates. Figure. S6 Flow cytometric analysis of pHexR-GFP with expression of gnd and/or gntR genes at 9 hours cultivation. Fluorescence histograms of pHexR-GFP ALE-1, ALE-1 pGnd, and ALE-1 pGnd-GntR strain were compared. Gate M1 was used to differentiate the population of wild-type and EMPP-deficient strains based on GFP intensity. Figure. S7 Diverse phenotypic characteristics of pfk_ALE-1. (a, b) Lycopene production was examined in ALE-1 and control strains all harboring the pAC-LYCO04 plasmid. (a) Lycopene titer of E. coli strains after 24 h of cultivation were compared. (b) Color of lycopene-producing strains after 24 h of cultivation. Two milliliters of lycopene-producing cells were harvested after 24 h of cultivation and re-suspended in 200 µL of distilled water. After transfer to 96-well plates, images of cells were taken using a 96-well microplate reader. (c) When nitrate was used as the final electron acceptor under anoxic conditions, OD600 of ALE-1 and MG1655 were measured. For OD600 of MG1655 with nitrate addition, there was no significant difference caused by nitrate addition (data not shown). (d–f) Residual sugar concentration and cell growth of ALE-1 and MG1655 strains when grown in M9 minimal media supplemented with 3 g/L of glucose and 2 g/L of xylose. Residual glucose and xylose concentration of (d) MG1655 and (e) ALE-1 strain. (f) The cell growth of both strains is presented. Figure. S8 Unexpected expression level of the pfkB gene of ALE-1. In the process of deletion of the pfkB gene, nucleotides of 157bp from the 5′ end and 177bp from the 3′ end were detected. Using bowtie software, the sequence reads were mapped onto the wild-type reference genome (NC_000913.2) with the maximum insert size option of 500bp. Remained coding sequence of the pfkB gene affected the expression level of the pfkB gene of ALE-1. Table S1. Strains and plasmid list. Table S2. Primer List.

In bacterial transcription, transcription initiation is arguably the most important regulatory point, because transcribing unnecessary genes into RNA could be a waste of energy, time and resources. There are multiple components which are involved in bacterial transcription initiation: RNA polymerase, [sigma]-factors, transcription factors, and transcription start sites. Each component has been intensively investigated, however in a limited scope and mostly with low-throughput methods. New technologies, such as hybridization on microarray and deep-sequencing, enabled researchers to study each component in a systems level, in a combination of two or more components, and in comparison between different species. In order to facilitate the analysis, integration, and comparison, software, MetaScope, was developed to accommodate multiple genome-scale datasets to visualize, analyze, integrate, and compare. TSS-seq, modified 5'-RACE with deep- sequencing, gave a genome-scale landscape of transcription start sites, and comparison of TSSs of conserved genes between closely-related species, E. coli and K. pneumoniae, showed significantly different usage of promoters, which implies different regulation of orthologous genes. To further investigate properties of promoters which were identified by TSS-seq, ChIP-chip experiments were performed for [sigma]-factors in E. coli to determine [sigma]-factor regulons. From the reconstructed [sigma]- factor network, extensive overlaps between regulons were observed. [sigma]⁷⁰ and [sigma]³⁸ share the largest set of genes in E. coli, and additional experiments revealed that those [sigma]-factors work in competition and utilize the negative regulation by [sigma]³⁸ . ChIP-exo, which applies exonuclease to present better resolution of DNA-binding, and RNA-seq implemented more detailed identification of Fur regulon in E. coli. Reconstruction of Fur regulon completed the previous knowledge of bacterial response to iron change, and also enabled its role over iron metabolism. In order to understand how bacteria respond to nitrogen limitation, the same methods were used under conditions that were predicted from model-based prediction, and resulted in reconstruction of regulons for major transcription factors, NtrC and Nac. Determination of those regulons expanded the current knowledge of nitrogen metabolism and how it is regulated in bacteria. Thus, systems approaches enabled a genome-scale assessment of regulatory components in multiple levels, and contributed to expansion of the current knowledge of bacterial transcription initiation