Journal: Nucleic Acids Research

Article Title: Dissecting host stress responses for predictable heterologous gene expression in E. coli

doi: 10.1093/nar/gkag256

Figure Lengend Snippet: Workflow for identifying cellular stress responses to heterologous gene expression. ( a ) Experimental workflow. Expression plasmids were transformed into E. coli and cultivated at 37°C. RNA-seq samples were collected following induction, here illustrated as a marked induction timepoint on a growth curve. iModulon signals were extracted from transcriptomes using ICA, enabling characterization of transcriptional, translational, and product-specific stress contributions, each illustrated as cell icons. ( b ) Design of protein library. The library consisted of 12 different heterologous proteins spanning 0–50 kDa with varying cysteine (0%–25%) and tyrosine (0%–25%) content. Proteins include whey/egg white, eGFP and eGFP fusions, cysteine-rich proteins, tyrosine-rich protein (MFP5), and proteins that fold well (MBP, MNEI). ( c ) ICA methodology. ICA decomposes a gene expression matrix ( X ) into gene weights ( M ) and iModulon activities ( A ), represented as blue, gray, and green rectangles. Gene weights define the membership of genes to each iModulon and relate gene expression to iModulon activities gene expression to iModulon activities across samples ( X = M*A). RNA-seq dataset from this study ( N = 74) was combined with the PRECISE-1K dataset ( N = 1035) to calculate 156 iModulons from 4257 genes.

Article Snippet: We plotted RpoH versus Proteostasis iModulon activities across all (>1000) PRECISE-1K transcriptomes (Fig. ).

Techniques: Gene Expression, Expressing, Transformation Assay, RNA Sequencing

Journal: Nucleic Acids Research

Article Title: Dissecting host stress responses for predictable heterologous gene expression in E. coli

doi: 10.1093/nar/gkag256

Figure Lengend Snippet: Hierarchical clustering reveals distinct transcriptional stress profiles. ( a ) Dendrogram of hierarchical clustering based on transcriptome similarity across all expression conditions. Each condition is annotated with the expressed protein, RNAP system, RBS variant, additional details (Info), heterologous mRNA% (percent of total transcripts), and cell density at RNA-seq sampling (OD 600 ). The latter two are here visualized as horizontal bar charts overlaid with individual replicates. Samples were collected 2 h after induction at OD 600 = 0.40. Samples marked with an asterisk were taken 3 h after induction. Five distinct clusters were identified: Control (gray), mRNA Stress (blue), mRNA + Protein Stress (purple), Protein Stress #1 (light green), and Protein Stress #2 (dark green). ( b ) iModulon activity signatures for each cluster. Strip plots show activities (scale: −40 to 40) of the top eight differentially activated iModulons per cluster; individual points represent single samples.

Article Snippet: We plotted RpoH versus Proteostasis iModulon activities across all (>1000) PRECISE-1K transcriptomes (Fig. ).

Techniques: Expressing, Variant Assay, RNA Sequencing, Sampling, Control, Activity Assay, Stripping Membranes

Journal: Nucleic Acids Research

Article Title: Dissecting host stress responses for predictable heterologous gene expression in E. coli

doi: 10.1093/nar/gkag256

Figure Lengend Snippet: Characterizing protein production stress responses and engineering strain and media improvements. ( a ) Scatterplots showing the correlation between eGFP protein levels (median FITC-A) and iModulon activities for RpoH (top) and Proteostasis (bottom). Pearson correlation coefficients ( r ) and P -values are shown; shaded regions indicate 95% confidence intervals. ( b ) Gene membership comparison between RpoH and Proteostasis iModulons. Gene weights are plotted for each iModulon ( x -axis: RpoH; y -axis: Proteostasis). Dashed lines indicate membership thresholds. Labeled genes exceed membership thresholds in their respective iModulons (green: RpoH genes; blue: Proteostasis genes). ( c ) Function of RpoH and Proteostasis iModulon genes. RpoH genes provide chaperone activity, disaggregation, proteolysis, and refolding. Proteostasis genes regulate osmotic stress and stationary phase adaptation. ( d ) Phase-plane plot comparing RpoH ( x -axis) and Proteostasis ( y -axis) iModulon activities across PRECISE-1K samples ( n > 1000; see “Materials and methods” section). Samples from heat shock experiments (yellow: 30°C; orange: 37°C; red: 44°C) and Protein Stress clusters #1 and #2 (green) are highlighted; remaining samples are gray. ( e ) Violin plots comparing mRNA levels (log-TPM) of rpoH, ibpA , and ibpB across control, protein production, and heat shock conditions. Statistical comparisons are shown with P -values (Student’s t -test). ( f ) Effect of rspAB modulation on eGFP protein production. Top schematics show design of base strain, rspAB overexpression (OE), and rspAB knockout (KO). Time courses show development of eGFP production (mean FITC-A, circles, left y -axis) and cell density (OD 600 , crosses, right y -axis). ( g ) Effect of media supplementation with osmolytes (5 mM choline or 5 mM betaine) in eGFP-producing strains. Mean percent changes (Δ) and P -values (Student’s t -test) are used in panels ( f ) and ( g ) to compare eGFP levels from final timepoint samples to base controls.

Article Snippet: We plotted RpoH versus Proteostasis iModulon activities across all (>1000) PRECISE-1K transcriptomes (Fig. ).

Techniques: Comparison, Labeling, Activity Assay, Control, Over Expression, Knock-Out

Journal: Nucleic Acids Research

Article Title: Dissecting host stress responses for predictable heterologous gene expression in E. coli

doi: 10.1093/nar/gkag256

Figure Lengend Snippet: Product-specific cellular stress responses. ( a ) Cold Shock iModulon activity ( y -axis) versus heterologous mRNA ( x -axis, % of total transcripts). Colored ellipses represent the five main transcriptome clusters (identified in Fig. ). An arrow indicates the trend of the Cold Shock response to increasing heterologous mRNA levels. Labeled points indicate individual expression conditions outside these clusters. ( b ) Proteostasis iModulon activity ( y -axis) versus RpoH iModulon activity ( x -axis). Colored ellipses represent the five main transcriptome clusters. An arrow indicates the direction of RpoH and Proteostasis iModulon activities to increasing protein stress. Labeled points show individual conditions. ( c ) Left: Horizontal bar chart with protein lengths and cysteine content (%) for all expressed proteins, ranked by cysteine percentage. Right: Heatmap of average iModulon activities across expression conditions with activities displayed numerically in each cell. These iModulons are related to iron regulation (Fur-1, Fur-2), Fe-S cluster assembly (Isc System, Suf System), oxidative stress (SoxS, NO Stress, OxyR), NrdR, thiamine biosynthesis, and transposon activity (IS Elements-4).

Article Snippet: We plotted RpoH versus Proteostasis iModulon activities across all (>1000) PRECISE-1K transcriptomes (Fig. ).

Techniques: Activity Assay, Labeling, Expressing