RNase E cleavage shapes the transcriptome of Rhodobacter sphaeroides and strongly impacts phototrophic growth

Abstract

Bacteria adapt to changing environmental conditions by rapid changes in their transcriptome. This is achieved not only by adjusting rates of transcription but also by processing and degradation of RNAs. We applied TIER-Seq (transiently inactivating an endoribonuclease followed by RNA-Seq) for the transcriptome-wide identification of RNase E cleavage sites and of 5' RNA ends, which are enriched when RNase E activity is reduced in Rhodobacter sphaeroides. These results reveal the importance of RNase E for the maturation and turnover of mRNAs, rRNAs, and sRNAs in this guanine-cytosine-rich α-proteobacterium, some of the latter have well-described functions in the oxidative stress response. In agreement with this, a role of RNase E in the oxidative stress response is demonstrated. A remarkably strong phenotype of a mutant with reduced RNase E activity was observed regarding the formation of photosynthetic complexes and phototrophic growth, whereas there was no effect on chemotrophic growth.

Figure 1.. Generation of stable 5′ ends in wild type and 2.4.1rne^{E. coli(ts)} strains of R. sphaeroides.

(A) Internal cleavage by RNase E. The ends of primary transcripts are protected from degradation by a triphosphate at their 5′ end and by secondary structures (often terminators) at their 3′ end. Internal cleavage of RNase E generates an unprotected 3′ end, which allows rapid degradation by 3′–5′ exoribonucleases. The new monophosphorylated 5′ end is a new target for RNase E. The 5′ end stemming from RNase E cleavage is accumulated in the wild type at the nonpermissive temperature. (B) 5′ end–dependent degradation by RNase E. RNase E can bind to monophosphorylated 5′ ends if a stretch of unpaired nt is present and will subsequently introduce cleavages in an overall 5′–3′ direction. The 5′ monophosphate ends can stem from previous RNase E cleavage, from cleavage by other endoribonucleases, or by the action of a pyrophosphohydrolase. The 5′ monophosphate end is enriched in the rne mutant strain compared with the wild type at the nonpermissive temperature. Global analysis of 5′ end profiles at a permissive (32°C) (C) and a nonpermissive temperature (42°C) (D). The plots show average first-base-in-read-coverage level in wild-type samples compared with 2.4.1rne^{E. coli(ts)} samples and the relative log₂ fold change. The x axis (base mean) represents the average, library-size normalized coverage values, whereas the y axis (fold change) represents the ratio of the normalized coverage values of the mutant in comparison with the wild type (both base-means and fold changes were calculated by DESeq2). The green dots represent sites with a significant (i.e., Benjamini–Hochberg adjusted, P < 0.05) enrichment in the wild type, whereas the brown dots represent sites with a significant enrichment in the mutant. Black dots represent sites without a significant enrichment.

Figure S1.. Scatterplot of PCA done as part of the DESeq2 analysis.

Figure 2.. Statistics of cleavage sites and mutant-enriched sites.

(A) Sum of all RNase E cleavage sites mapped to different RNA categories. (B) Sum of all enriched sites mapped to different RNA categories. (C) Box plot of the cleavage site density for the individual genes per kilobase sorted by gene type. (D) Box plot of mutant enriched site density for the individual genes per kilobase sorted by gene type. (E) Histogram of the distribution of RNase E cleavage site density in genes per kilobase. (F) Histogram of the distribution of mutant enriched site density in genes per kilobase.

Figure 3.. Consensus sequence motifs.

(A) Sequence motif of RNase E cleavage sites based on alignment of all mapped cleavage sites (with one base of shifting allowed). (B) Sequence motif around 5′ ends enriched in the rne mutant based on the alignment of all mapped cleavage sites (with one base of shifting allowed).

Figure S2.. Frequency of cleavage site dependent on their relative position (A) to start codon and (B) to stop codon.

Figure 4.. Northern blot analysis of the sRNAs UpsM (A) and SorX (B) in the wild type and the 2.4.1rne^{E. coli(ts)} mutant.

For RNA isolation, the strains were cultivated under microaerobic conditions at 32°C and shifted for 20 min to 42°C. 10 μg of total RNA was blotted. 5S rRNA was used as a loading control. Read coverage for UpsM (A) and SorX (B) in the wild type (green/light green) and 2.4.1rne^{E. coli(ts)} (brown/orange) at 32°C (dark colors) and 42°C (bright colors). Upper panels show the coverage of all reads, whereas the lower panels show the 5′ ends with single-nucleotide resolution.

Figure S3.. Northern blot analysis of the sRNAs 1624 (A), 1771 (B), SorY (C), and PcrZ (D), and the mRNAs RSP_7527 (E) and RSP_7517 (F) in the wild type and the 2.4.1rne^{E. coli(ts)} mutant.

Strains for RNA isolation were cultivated under microaerobic conditions at 32°C and shifted to 42°C for 20 min. 10 μg of total RNA was plotted. 5S rRNA probe was used as loading control. Read coverage for the wild type is shown in green/light green and for 2.4.1rne^{E. coli(ts)} mutant, in brown/bright orange at 32°C (dark colors) and 42°C (bright colors). Upper panels show the coverage of all reads, whereas the lower panels show the 5′ ends with single-nucleotide resolution.

Figure S4.. Bacteriochlorophyll a concentration in the wild-type and the 2.4.1rne^{E. coli(ts)} strain.

Cells for BChl measurement were cultivated under either microaerobic (LO) or phototrophic conditions (PT) at 32°C (n = 3).

Figure 5.

(A) Absorption spectra of cell-free extracts. Spectra were obtained from cultures grown either under microaerobic or phototrophic conditions to an OD₆₆₀ of 0.8. The absorption between 450 and 600 nm is caused by carotenoids, whereas the peaks at 800 and 850 nm are caused by the BChls bound to LHII (B800-850) and LHI (B870). The spectra are each for one representative culture. (B) Cultures of the wild type of R. sphaeroides (green) and 2.4.1rne^{E. coli(ts)} (orange) mutant were cultivated at 32°C under microaerobic conditions in the dark or in the presence of light (60 W) under anaerobic conditions. Optical density was measured at 660 nm. The curves represent the mean value of biological triplicates. (C) Survival assay of the wild type of R. sphaeroides and 2.4.1rne^{E. coli(ts)} mutant. The strains were cultivated at 32°C under microaerobic conditions in the dark to an OD₆₆₀ of 0.4 and shifted to 42°C for 1 h either under no-stress conditions for control or in the presence of either 0.5 mM H₂O₂, 1.5 mM tBOOH, or 1 mM paraquat. For photooxidative stress, bacteria were cultivated at 32°C under aerobic conditions in the dark and in the presence of 0.2 μM methylene blue until reaching an OD₆₆₀ of 0.4. The cells were then shifted to 42°C in the presence of light (800 W/m²) for 1 h, whereas the controls were cultivated in the dark. Survival is displayed as the percentage of colony numbers forming in serial dilutions of the stressed bacterial cultures relative to the control culture (referred to as 100% survival). The average of two measurements of two independent cultures is shown.

Figure S5.. pufL coverages.

Figure S6.. Zone of inhibition assay.

R. sphaeroides strains were cultivated under microaerobic conditions at 32°C in biological triplicates. 200 μl culture with an OD₆₆₀ of 0.4 was mixed with prewarmed top agar (0.8% agar) and layered onto malate minimal salt medium plates. When top agar was solidified, a filter disk containing 5 μl of the oxidative stress agent was placed in the center of the plate. Concentration of the agents used are 1 M H₂O₂, 0.5 M tBOOH, 0.5 M paraquat, and 10 mM methylene blue (MB). Values of the diameter (millimeter) of the zone of growth inhibition are displayed in the diagram. The given values are the means of the technical duplicates of three independent biological replicates.

Figure 6.. Schematic overview of the importance of RNase E for R. sphaeroides.

Although RNase E is only one of several RNases acting on mRNAs and sRNAs, it significantly affects the transcriptome of R. sphaeroides, which most likely also affects the proteome. These effects on RNA processing/degradation strongly impact phototrophic growth, whereas no significant effect on chemotrophic growth is observed.