Comparative analyses of the variation of the transcriptome and proteome of Rhodobacter sphaeroides throughout growth

Abstract

Background: In natural environments, bacteria must frequently cope with extremely scarce nutrients. Most studies focus on bacterial growth in nutrient replete conditions, while less is known about the stationary phase. Here, we are interested in global gene expression throughout all growth phases, including the adjustment to deep stationary phase.

Results: We monitored both the transcriptome and the proteome in cultures of the alphaproteobacterium Rhodobacter sphaeroides, beginning with the transition to stationary phase and at different points of the stationary phase and finally during exit from stationary phase (outgrowth) following dilution with fresh medium. Correlation between the transcriptomic and proteomic changes was very low throughout the growth phases. Surprisingly, even in deep stationary phase, the abundance of many proteins continued to adjust, while the transcriptome analysis revealed fewer adjustments. This pattern was reversed during the first 90 min of outgrowth, although this depended upon the duration of the stationary phase. We provide a detailed analysis of proteomic changes based on the clustering of orthologous groups (COGs), and compare these with the transcriptome.

Conclusions: The low correlation between transcriptome and proteome supports the view that post-transcriptional processes play a major role in the adaptation to growth conditions. Our data revealed that many proteins with functions in transcription, energy production and conversion and the metabolism and transport of amino acids, carbohydrates, lipids, and secondary metabolites continually increased in deep stationary phase. Based on these findings, we conclude that the bacterium responds to sudden changes in environmental conditions by a radical and rapid reprogramming of the transcriptome in the first 90 min, while the proteome changes were modest. In response to gradually deteriorating conditions, however, the transcriptome remains mostly at a steady state while the bacterium continues to adjust its proteome. Even long after the population has entered stationary phase, cells are still actively adjusting their proteomes.

Fig. 1

Changes of RNA levels as monitored by microarrays throughout the growth phases. Levels of RNAs in later growth phases were normalized to RNA levels in exponential phase. Box plots (a) give an overview on the transcriptome variability of the different samples. Kernel density estimators (KDE) (b) compare the distribution of growth phase-dependent changes in gene expression. The heat map (c) gives a global overview of the changes in RNA levels at single gene resolution. The ranking of genes is based upon the degree of log₂FC observed in transition phase, with the highest increases in transcript abundance in the transition phase appearing at the top of the heat map. Indicated underneath the heat map is the total number of transcripts detected at each growth phase

Fig. 2

Changes of protein levels as monitored by quantitative mass spectrometry throughout the growth phases. Levels of proteins in later growth phases were normalized to protein levels in the exponential phase. Box plots (a) give an overview on the proteome variability of the different samples. Kernel density estimators (KDE) (b) compare the distribution of growth phase-dependent changes in protein levels. The heat map (c) gives a global overview of the changes in protein levels for each individual gene. The ranking of genes is based upon the degree of log₂FC observed in transition phase, with the highest increases in protein abundance in the transition phase appearing at the top of the heat map. Indicated underneath the heat map is the total number of proteins detected at each growth phase

Fig. 3

Percentage of transcripts (a) and proteins (b) with changed levels throughout growth. For both the microarray and proteome data, the status of changed versus not-changed were determined using thresholds of log₂FC 0.65 and − 0.65. In each case, the percentage of transcripts or proteins with levels exceeding the cutoffs were determined for each growth phase. For a direct comparison, both graphs harbor identical scales on the ordinate

Fig. 4

Number of proteins with decreased levels (represented as negative values on the Y-axis) or increased levels (represented as positive values on the Y-axis) throughout growth in selected COGs, determined using thresholds of log₂FC 0.65 and − 0.65. The complete data set is provided in Additional file 4: Table S4