Sequence features of E. coli mRNAs affect their degradation

Abstract

Degradation of mRNA in bacteria is a regulatory mechanism, providing an efficient way to fine-tune protein abundance in response to environmental changes. While the mechanisms responsible for initiation and subsequent propagation of mRNA degradation are well studied, the mRNA features that affect its stability are yet to be elucidated. We calculated three properties for each mRNA in the E. coli transcriptome: G+C content, tRNA adaptation index (tAI) and folding energy. Each of these properties were then correlated with the experimental transcript half life measured for each transcript and detected significant correlations. A sliding window analysis identified the regions that displayed the maximal signal. The correlation between transcript half life and both G+C content and folding energy was strongest at the 5' termini of the mRNAs. Partial correlations showed that each of the parameters contributes separately to mRNA half life. Notably, mRNAs of recently-acquired genes in the E. coli genome, which have a distinct nucleotide composition, tend to be highly stable. This high stability may aid the evolutionary fixation of horizontally acquired genes.

Figure 1. A sliding window analysis of the correlation between the E. coli transcript G+C content and its respective half-life in the two data sets.

The strongest correlated window starts at nucleotide -40 (Spearman R = −0.185, p = 0.002 and R = −0.16 p = 0.003 respectively).

Figure 2. Average half lives of E. coli transcripts binned according to G+C content.

The number of genes (N) is listed above each group's designation. Vertical bars represent 95% confidence intervals. Boxes represent the mean of the group. A- M9 media; B- LB media.

Figure 3. A sliding window analysis of the correlation between the folding energy of E. coli transcripts and their respective half-life in each data set.

The strongest correlated window starts at nucleotide -5 (Spearman R = 0.150, p = 0.00015 and R = 0.151, p = 0.00014 respectively).

Figure 4. Effect of tAI on transcript stability.

The normalized E. coli dataset was divided into 3 subsets of genes according to their tAI. The number of genes (N) is listed above each group's designation. Vertical bars represent 95% confidence intervals. Boxes represent the mean half life of the group. A- M9 media; B- LB media.

Figure 5. Difference in transcript stability between foreign and native genes.

The E. coli genome was divided into two subsets of genes: recently transferred genes (LGT) (A: Genes described as recent LGT in , B: Genes described as LGT in , B) and non-recently transferred genes (Native). The number of genes (N) is listed above each group's designation. Vertical bars represent 95% confidence intervals. Boxes represent the mean of the group.