Figure 1.
Design of a comparative “omics”-approach for bacterial stress responses.
Bacterial cultures are subjected to stress (e.g., singlet oxygen, 1O2) and relative changes monitored for mRNAs in total RNA isolates (transcriptome), mRNAs in polysomes (translatome), and for proteins (proteome). For the proteome approach, a culture has to be fully labeled with the heavy amino acid 13C6-lysine (Lys6; heavy standard). Protein samples from a reference (no stress) or stressed cells are separately mixed in a 1∶1 ratio with heavy standard protein and subsequently applied to protein digest followed by MS analysis. SILAC ratios are calculated by comparing intensities of heavy (red) to light (green) peaks of individual peptides (m, mass and z, charge). Direct ratios, reflecting relative changes, are determined thereof. The translatome is assessed by microarray analysis of RNA from polysome fractions. Polysomes are enriched by sucrose density gradient centrifugation of cells that have been treated with chloramphenicol. For transcriptome data, total RNA is isolated and applied to microarray analysis.
Table 1.
Number of quantified and regulated genes in individual approaches.
Figure 2.
Expression kinetics of stress-related mRNAs and sRNAs.
Relative changes of mRNAs upon singlet oxygen (1O2) stress were monitored by microarray analysis of total RNA at time-points 7, 45, and 90 min and depicted as log2 ratios. (A) All mRNAs with significant induction (log2 ratio ≥0.8 and p-value <0.05) at one of the three time-points were applied to clustering using MeV (Multi Experiment Viewer version 4.7.4) from the TM4 Microarray Software Suite [67], [68]. Clustering was based on k-means (KMC method) according to Euclidean distance with a maximum of 50 iterations. Changes were illustrated as heat-maps with a color code ranging from −0.5 (green) to 1.5 (red) log2 ratio. Columns represent time-points of the stress response and rows represent individual genes. (B–F) Up-regulated mRNAs shown in (A) were grouped together according to their function. Expression kinetics, representing the mean of log2 ratios, were calculated for functional groups of cluster 1 (B), cluster 2a (C), cluster 2b (D), cluster 2c (E), and cluster 3 (F). (G) Analysis of stress-induced sRNAs. Total RNA was isolated at the indicated time-points and applied to Northern blot analysis. RSs0019, RSs0680a, and RSs0827 were hybridized to radioactively labeled oligonucleotides and visualized by phosphoimaging. 5S rRNA was probed as a control for equal loading of samples (not shown). (H) Genes with described RpoE- and RpoH2-dependent promoters [28], [29], [72] were analyzed when they exhibited log2 ratios ≥0.8. The curves represent the mean of log2 ratios based on 12 mRNAs (RpoE, black line) and 42 mRNAs (RpoH2, red line). See supplementary Tables S1 and S2 for further information on regulated genes and their particular functions.
Figure 3.
Accurate quantification of protein changes by SILAC.
Relative changes in protein abundance after 90 min of singlet oxygen stress were determined by an indirect quantification approach using a heavy standard labeled with 13C6-lysine (Lys6). Protein mixtures were digested and used for MS analysis (see Materials and Methods). In the central volcano plot the direct ratios (log2) of 1214 quantified proteins of quadruplicates were plotted against negative logarithmized p-values (log10). The histogram on the top shows log2 protein ratio distributions (Gaussian distribution). Up- and down-regulated proteins were grouped according to their functions which relate to stress defense (red triangles), proteases (black triangles), redox reactions (dark blue triangles), carbohydrate metabolism (grey triangles), transport processes (light blue triangles), photosynthesis (green triangles), and motility/chemotaxis (purple triangles). See supplementary Table S3 for further information on regulated proteins and their particular functions.
Figure 4.
RNA-seq validation for ORFs newly identified by SILAC.
ORFs detected by SILAC-based MS were compared to deep sequencing data (RNA-seq). (A–H) Integrated Genome Browser (Affymetrix) screenshots depict gene loci with nucleotide positions referring to the plus-strand. RNA-seq data are visualized as numbers of cDNA reads per nucleotide (red plots). Grey and blue arrows represent annotated and new ORFs, respectively. For further details on new ORFs see Table S4.
Figure 5.
Correlation between global approaches.
Scatter-plots represent pairwise comparisons of log2 ratios between (A) transcriptome (Total 90/0 min) and translatome (Poly 90/0 min), (B) proteome (SILAC 90/0 min) and transcriptome, (C) proteome and translatome. Number (n), Pearson correlation (r), and squared correlation (r2) of shared features are given for every comparison. It is indicated for up-regulated (log2 ratio ≥0.8) and down-regulated (log2 ratio ≤−0.8) features whether changes are unidirectional (log2 ratio difference between approaches <0.4, green spots) or biased (log2 ratio difference ≥0.4, red spots). (D) Functional grouping of genes that were newly identified to be stress-responsive by the integrative approach. For a complete list of genes and information on their function see Table S5.
Figure 6.
Polarity of stress-induced operons.
Graphs representing relative changes for selected operons after 90 min of singlet oxygen stress. Log2 ratios were derived from transcriptomic (Total 90 min, black bars), translatomic (Poly 90 min, red bars), and proteomic (SILAC 90 min, blue bars) data sets. Orientation of genes within operons is depicted from the left to the right, irrespective of their location in the genome, with the leftmost gene representing the first gene in the operon. Gene numbers refer to corresponding RSP-numbers. Operons were selected according to following categories: transport process and iron metabolism (A), RpoE-dependent (B, C, D), stress defense (C), transport process (D), chaperone (E), RpoH2-dependent (F, G), carbohydrate metabolism (F), redox reaction (G), protease (H).