Toxin-antitoxin genes are differentially expressed in Escherichia coli relA and spoT mutans cultured under nitrogen, fatty acid, or carbon starvation conditions

Abstract

Introduction: The stringent response is one of the fundamental mechanisms that control and modulate bacterial adaptation to stress conditions, such as nutrient limitation. The accumulation of stringent response effectors, (p)ppGpp, causes differential expression of approximately 500 genes, including genes of bacterial endogenous toxin-antitoxin (TA) systems. However, the exact link between (p)ppGpp and toxin-antitoxin systems' activation, as well as toxin-antitoxin role in stress adaptation remains disputed.

Methods: In this study, we performed a complex analysis of changes (RNA-Seq) in the toxin-antitoxin operons' transcription in response to nitrogen, fatty acid, or carbon starvation, in bacteria with different abilities of (p)ppGpp accumulation.

Results and discussion: Although we observed that in some cases (p)ppGpp accumulation appears to be crucial for transcriptional activation of TA genes (e.g., ghoST, ryeA), our data indicates that the general pattern of chromosomally encoded TA gene expression in E. coli differs depending on the nutrient distribution in the environment, regardless of the alarmone accumulation.

Keywords: RNA-seq; ghoST; hokB; mqsRA; ppGpp; relA; spoT203; toxin-antitoxin.

Figure 1

Overview of gene expression changes in ΔrelA, ΔrelA ΔspoT, and ΔrelA spoT203 mutants upon nutritional stress. Stress conditions: nitrogen (A), fatty acid deprivation (B), and carbohydrate deprivation (NH⁴⁺) deprivation (C) compared to no stress (normal). Volcano plots illustrate log2 fold change (x-axis) and statistical significance of changes (y-axis) as calculated by DESeq2 package. Horizontal dashed red line indicates p value corresponding to adjusted p value = 0.05. Purple vertical lines indicate casual log2FC = 2 threshold and red dashed vertical lines indicate MAD thresholds calculated for each comparison. Genes which are significantly up or downregulated (|log2FC| >2 and adj. p < 0.05) are colored in pink and blue respectively, whereas genes colored in gray are not significant. TA genes are indicated by large points. Labels are shown only for significantly altered TA genes.

Figure 2

Changes in Toxin and Antitoxin Transcript Levels Due to Nutritional Stress. The heatmap illustrates the log2 fold change (log2FC) in transcript levels of toxin-antitoxin (TA) systems in cells subjected to various nutritional stresses for 1.5 generations, as assessed by Illumina NGS RNA-seq. Stress conditions: carbohydrate deprivation (Carb.), fatty acid deprivation (FA), and nitrogen (NH⁴⁺) deprivation (Nitro.) compared to no stress (normal). Mutant Strains: ΔrelA; ΔrelA, ΔspoT (ppGpp null); ΔrelA, spoT203 (ppGpp high). Top Sidebar: Indicates the stress condition and mutant strain and left Sidebar: Indicates TA type and gene function (antitoxin or toxin). Color scale represents the exact log2FC values for the comparison of stress vs. normal conditions. Log2FC values for genes, mutants and conditions are provided in cells. Asterisk indicates statistical significance evaluated by DESeq2, adjusted p < 0.05. Small RNAs ralA and timA were not annotated in the genome version used in this study thus data for them were not available (n.a.).

Figure 3

Illustration of the possible analogy between (A) the generally accepted mechanism of bacteria killing by plasmid-encoded toxin activation due to post-segregational killing (PSK) mechanism and (B) differential expression of chromosomally encoded TA systems. Created with Biorender.com.