Global gene expression during stringent response in Corynebacterium glutamicum in presence and absence of the rel gene encoding (p)ppGpp synthase

Abstract

Background: The stringent response is the initial reaction of microorganisms to nutritional stress. During stringent response the small nucleotides (p)ppGpp act as global regulators and reprogram bacterial transcription. In this work, the genetic network controlled by the stringent response was characterized in the amino acid-producing Corynebacterium glutamicum.

Results: The transcriptome of a C. glutamicum rel gene deletion mutant, unable to synthesize (p)ppGpp and to induce the stringent response, was compared with that of its rel-proficient parent strain by microarray analysis. A total of 357 genes were found to be transcribed differentially in the rel-deficient mutant strain. In a second experiment, the stringent response was induced by addition of DL-serine hydroxamate (SHX) in early exponential growth phase. The time point of the maximal effect on transcription was determined by real-time RT-PCR using the histidine and serine biosynthetic genes. Transcription of all of these genes reached a maximum at 10 minutes after SHX addition. Microarray experiments were performed comparing the transcriptomes of SHX-induced cultures of the rel-proficient strain and the rel mutant. The differentially expressed genes were grouped into three classes. Class A comprises genes which are differentially regulated only in the presence of an intact rel gene. This class includes the non-essential sigma factor gene sigB which was upregulated and a large number of genes involved in nitrogen metabolism which were downregulated. Class B comprises genes which were differentially regulated in response to SHX in both strains, independent of the rel gene. A large number of genes encoding ribosomal proteins fall into this class, all being downregulated. Class C comprises genes which were differentially regulated in response to SHX only in the rel mutant. This class includes genes encoding putative stress proteins and global transcriptional regulators that might be responsible for the complex transcriptional patterns detected in the rel mutant when compared directly with its rel-proficient parent strain.

Conclusion: In C. glutamicum the stringent response enfolds a fast answer to an induced amino acid starvation on the transcriptome level. It also showed some significant differences to the transcriptional reactions occurring in Escherichia coli and Bacillus subtilis. Notable are the rel-dependent regulation of the nitrogen metabolism genes and the rel-independent regulation of the genes encoding ribosomal proteins.

Figure 1

The m/a scatter plot comparing the transcriptomes of the C. glutamicum rel-deficient and the rel-proficient strains. Genes with a reduced transcription are indicated by red dots and those with an upregulated transcription by green dots. Genes described in the text and showing the strongest changes in the expression ratio are marked by arrows.

Figure 2

Time course analysis of expression of the histidine and serine biosynthesis genes after treatment with DL-serine hydroxamate. The samples were taken in five minute intervals up to a total of 25 minutes. The relative amount of mRNA was measured by real-time RT-PCR with a LightCycler instrument. The expression ratios were determined by calculating the ratio of t₀ (untreated) and t_5–25 (treated) samples from four experiments. Below the diagrams, the organization of the respective genes in the C. glutamicum chromosome is depicted.

Figure 3

DNA microarray hybridizations of the C. glutamicum RES167 and the C. glutamicum RES167Δrel mutant strain. (A) and (B) Scatter plots (m/a plots) of the experiments. Shown are all genes whose transcription has been significantly up- (green) or downregulated (red) after induction of the stringent response by DL-serine hydroxamate. Genes which showed the most prominent changes in expression are labelled. (C) Venn diagram of the differentially expressed genes in the C. glutamicum RES 167 and Δrel mutant after treatment with SHX is shown. The numbers of genes which were placed in each of the three classes A, B and C is given. Green numbers represent upregulated genes, red ones represent downregulated genes.

Figure 4

Graphical representation of selected genes and putative operons comprising genes of classe A in their genomic organization. Below each gene, the induction or repression ratios are given for both strains, C. glutamicum RES167 and its derived Δrel mutant. Bold numbers indicate a significant induction or repression of the corresponding gene according to the filtering criteria (5% error probability) applied on the microarray data. Genes are not drawn to scale. Known transcription start points are marked by small arrows. Predicted rho-independent terminators are indicted by hairpins.

Figure 5

Illustration of putative operons of class B. For each experiment with the C. glutamicum RES 167 and its derived Δrel-mutant the induction or repression ratios are given. Bold numbers indicate a significant induction or repression of the corresponding gene according to the filtering criteria (5% error probability) applied on the microarray data. Genes are not drawn to scale. Known transcription start points are marked by small arrows. Predicted rho-independent terminators are indicted by hairpins.