Translational control plays an important role in the adaptive heat-shock response of Streptomyces coelicolor

Abstract

Stress-induced adaptations require multiple levels of regulation in all organisms to repair cellular damage. In the present study we evaluated the genome-wide transcriptional and translational changes following heat stress exposure in the soil-dwelling model actinomycete bacterium, Streptomyces coelicolor. The combined analysis revealed an unprecedented level of translational control of gene expression, deduced through polysome profiling, in addition to transcriptional changes. Our data show little correlation between the transcriptome and 'translatome'; while an obvious downward trend in genome wide transcription was observed, polysome associated transcripts following heat-shock showed an opposite upward trend. A handful of key protein players, including the major molecular chaperones and proteases were highly induced at both the transcriptional and translational level following heat-shock, a phenomenon known as 'potentiation'. Many other transcripts encoding cold-shock proteins, ABC-transporter systems, multiple transcription factors were more highly polysome-associated following heat stress; interestingly, these protein families were not induced at the transcriptional level and therefore were not previously identified as part of the stress response. Thus, stress coping mechanisms at the level of gene expression in this bacterium go well beyond the induction of a relatively small number of molecular chaperones and proteases in order to ensure cellular survival at non-physiological temperatures.

Figure 1.

Representative polysome profiles of S. coelicolor mycelial extracts from surface grown mycelium cultured at 30°C (A) and heat-shocked at 42°C (B) for 15 min. The mRNAs associated with monosome and polysome containing fractions were pooled and analysed for mRNA content separately.

Figure 2.

Changes in the average abundance of total mRNA and polysome/monosome associated transcripts following heat shock. Relative log₂ fold-change in transcript abundance (average of 5 biological replicates) between 42°C versus 30°C in total mRNA (the transcriptome; top panel), mRNA associated with polysome (middle panel) and mRNA associated with monosome fractions (bottom panel). In each panel, transcripts significantly up-regulated (PFP ≤ 0.1) at the higher temperature are highlighted in red, while those significantly down-regulated (PFP ≤ 0.1) are coloured blue. Non-significant transcripts are colored light grey, and all transcripts are ordered by their location on the chromosome.

Figure 3.

Changes in translational efficiency (TE) are not dependent on transcript abundance (T). (A) Identification of transcripts that exhibit a significantly altered translational efficiency (PFP < 0.1) following heat shock treatment, based on an analysis of the changes in average abundance of each transcript in the polysome relative to its abundance in the transcriptome. Transcripts that are more efficiently translated at 42°C compared to 30°C are indicated in red, while those showing less efficient translation are shown in blue. Non-significant transcripts are coloured light gray. (B) The significant changes in translational efficiency identified in A) (red and blue spots) arise from changes occurring at both the transcriptional and translational level. (C) Significant changes in transcript abundance in the transcriptome are not necessarily converted into corresponding changes in polysome association. Black or dark gray spots correspond to transcripts significantly up- or down-regulated (PFP ≤ 0.1), respectively, following heat shock but which exhibit no significant change in abundance in translational efficiency. Transcripts showing significant increases in both transcription and translational efficiency are shown in red, those exhibiting a significant decrease in transcription but with an increased translational efficiency are colored brown. The arrow indicates the location of the six cold-shock proteins in this group. Orange indicates those transcripts with significantly increased translational efficiency that do not significantly change in abundance in the transcriptome, and blue corresponds to the three transcripts with reduced translational efficiency from A) and B) that do not change significantly in the transcriptome. All non-significant transcripts are coloured light grey. Note that the same set of 74 genes is represented in Panel A, B and C. The four-way Venn diagram in Panel D represents the different subsets of genes that are significantly over-represented in the polysome pool, enhanced translational efficiency (TE), and upregulated or down-regulated at the transcriptional level (T); networks for some of these gene products are illustrated in Figure 4.

Figure 4.

Protein-protein interaction networks from the products of genes with enhanced translational efficiency (TE) following heat-shock. Proteins are shown for which there is a known or predicted interaction, derived from STRING 10.5 (33). Three different groups are highlighted, distinguished by whether transcription (T) is also signicantly altered. Those genes which are ‘potentiated’ (both translationally and transcriptionally induced) include the major molecular chaperone machines. Genes that are translationally induced with no significant change in transcription include universal stress reponse proteins and ABC transporters while genes that are translationally enhanced despite downregulation of transcription include the gene family encoding ‘cold-shock’ proteins. The key protein groups are colour coded and highlighted in boxes above the network.

Figure 5.

Transcripts that are more efficiently translated at 42°C compared to 30°C are distinguished by their nucleotide content and by the composition of their encoded proteins. (A). Proteins encoded by the more efficiently translated transcripts are significantly shorter than the genome population (left panel, compare Sig. to All; P-value from a one-tailed Wilcoxon test), and when considered as a set contain notably fewer amino acid residues than would be expected from the random sampling of 1000 sets of proteins of equal number (right panel, compare Sig. to Random. The dashed blue line is 2 standard deviations below the mean value for the random sets). The boxes plotted indicate the median, and the upper and lower quartile values of the distributions. (B) The amino acid composition of proteins encoded by the more efficiently translated transcripts (Sig.) is different to that observed by random sampling of 1000 sets of proteins of equal number (Random). Amino acids showing an increased frequency of use per 100 residues in the significant set are indicated by red boxes, and blue boxes mark less frequently used amino acids. Frequency values >2 standard deviations higher or lower in the significant set compared to the mean of the random sets are considered to be significantly different. (C) The GC nucleotide composition of the more efficiently translated transcripts (upper panel, red line) is reduced compared to that observed by random sampling of 1000 sets of proteins of equal number (upper panel, black line ± SD (shaded gray)). The average G+C composition across the genome is 72.1%. One-tailed t-tests indicate that the significant set possess significantly reduced G+C composition in the 50 bp region upstream of the translational start site, in the 50 bp of coding sequence downstream of the translational start site, and in the full coding sequence compared with an equivalent analysis of all genes (lower table). (D) Relative synonymous codon usage (RSCU) is altered in the genes encoding the more efficiently translated transcripts compared to that observed by random sampling of 1000 sets of proteins of equal number. Codons with RSCU values >2 standard deviations higher (red) or lower (blue) in the significant set compared to the mean of the random sets are considered to be significantly different.

Figure 6.

A subset of transcripts show significantly different changes in abundance in the monosome fractions relative to the polysome polysome fractions. (A) Identification of transcripts that exhibit a significantly altered partitioning between the monosome and polysome populations (PFP < 0.1) following heat-shock treatment, based on an analysis of the changes in abundance of each transcript in the monosome relative to its abundance in the polysome. Transcripts that become significantly more associated with the monosome fraction relative to the the polysome at 42°C compared to 30°C are indicated in red, while those showing a reduced association are shown in blue. Non-significant transcripts are coloured light gray. (B) Association of the significant changes identified in A) (red and blue points) with ORF lengths shorter than 1000 nucleotides (nts). (C) The significant changes identified in A) (red and blue points) correspond to transcripts with lower than average abundance at both 30°C and 42°C.