Limits...
Uncovering genes with divergent mRNA-protein dynamics in Streptomyces coelicolor.

Jayapal KP, Philp RJ, Kok YJ, Yap MG, Sherman DH, Griffin TJ, Hu WS - PLoS ONE (2008)

Bottom Line: Many biological processes are intrinsically dynamic, incurring profound changes at both molecular and physiological levels.Despite this overall correlation, by employing a systematic concordance analysis, we estimated that over 30% of the analyzed genes likely exhibited significantly divergent patterns, of which nearly one-third displayed even opposing trends.Our observations suggest that differences between mRNA and protein synthesis/degradation mechanisms are prominent in microbes while reaffirming the plausibility of such mechanisms acting in a concerted fashion at a protein complex or sub-pathway level.

View Article: PubMed Central - PubMed

Affiliation: Department of Chemical Engineering and Materials Science, University of Minnesota, Minneapolis, Minnesota, United States of America.

ABSTRACT
Many biological processes are intrinsically dynamic, incurring profound changes at both molecular and physiological levels. Systems analyses of such processes incorporating large-scale transcriptome or proteome profiling can be quite revealing. Although consistency between mRNA and proteins is often implicitly assumed in many studies, examples of divergent trends are frequently observed. Here, we present a comparative transcriptome and proteome analysis of growth and stationary phase adaptation in Streptomyces coelicolor, taking the time-dynamics of process into consideration. These processes are of immense interest in microbiology as they pertain to the physiological transformations eliciting biosynthesis of many naturally occurring therapeutic agents. A shotgun proteomics approach based on mass spectrometric analysis of isobaric stable isotope labeled peptides (iTRAQ) enabled identification and rapid quantification of approximately 14% of the theoretical proteome of S. coelicolor. Independent principal component analyses of this and DNA microarray-derived transcriptome data revealed that the prominent patterns in both protein and mRNA domains are surprisingly well correlated. Despite this overall correlation, by employing a systematic concordance analysis, we estimated that over 30% of the analyzed genes likely exhibited significantly divergent patterns, of which nearly one-third displayed even opposing trends. Integrating this data with biological information, we discovered that certain groups of functionally related genes exhibit mRNA-protein discordance in a similar fashion. Our observations suggest that differences between mRNA and protein synthesis/degradation mechanisms are prominent in microbes while reaffirming the plausibility of such mechanisms acting in a concerted fashion at a protein complex or sub-pathway level.

Show MeSH

Related in: MedlinePlus

Principal component analysis of transcriptome and proteome data.(A) ‘Loadings’ (eigenvector) plot for the first two principal component axes – PC-1 (solid lines) and PC-2 (dashed lines) from proteome (red lines) and transcriptome (blue lines) data (B) Percentage of variation accounted for by each of the seven principal components in proteome and transcriptome data (C) Values of genes along PC-1protein plotted against PC-1mRNA. Green and blue dots represent genes with significantly large difference in expression trends (/PC-1protein−PC-1mRNA/≥2). Of these, blue dots indicate those which are likely to exhibit opposing trends (2nd and 4th quadrants). All other genes are shown as red dots.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC2367054&req=5

pone-0002097-g004: Principal component analysis of transcriptome and proteome data.(A) ‘Loadings’ (eigenvector) plot for the first two principal component axes – PC-1 (solid lines) and PC-2 (dashed lines) from proteome (red lines) and transcriptome (blue lines) data (B) Percentage of variation accounted for by each of the seven principal components in proteome and transcriptome data (C) Values of genes along PC-1protein plotted against PC-1mRNA. Green and blue dots represent genes with significantly large difference in expression trends (/PC-1protein−PC-1mRNA/≥2). Of these, blue dots indicate those which are likely to exhibit opposing trends (2nd and 4th quadrants). All other genes are shown as red dots.

Mentions: We sought to identify the major trends in both mRNA and protein dynamics using Principal Component Analysis (PCA). Of the 894 genes selected earlier, we chose to analyze only 798 gene products for which quantification ratios from at least four out of eight time points could be determined from both microarray and proteomic experiments. Missing values were estimated by linear interpolation as before. Considering the eight time-point log transformed temporal gene expression profiles as data points in an eight-dimensional space, PCA enabled us to transform coordinate axes and identify the most important dimensions in this transformed space (dimensionality reduction thereby simplifying the time-course data). Independent PCA of both mRNA and protein data from the 798 gene products indicated that the first two principal components (PC-1mRNA and PC-1protein) account for over 85% of the total variance in the either dataset (Figure 4B). The loadings plots (eigenvectors) for each of these two principal components were remarkably well correlated between mRNA and protein data (Figure 4A) implying that the major trends in both proteome and transcriptome domains were strikingly similar. Furthermore, the principal component 1 represented the tendency of genes to be either up- or down-regulated during stationary phase adaptation. This component alone accounted for over 75% of the total variance in either dataset, and therefore, appears to be the most prominent pattern in our data. The other major component, PC-2, which accounted for an additional 6–9% variance, represents the tendency of certain genes to be transiently up- or down-regulated primarily during the transition between exponential and stationary phase.


Uncovering genes with divergent mRNA-protein dynamics in Streptomyces coelicolor.

Jayapal KP, Philp RJ, Kok YJ, Yap MG, Sherman DH, Griffin TJ, Hu WS - PLoS ONE (2008)

Principal component analysis of transcriptome and proteome data.(A) ‘Loadings’ (eigenvector) plot for the first two principal component axes – PC-1 (solid lines) and PC-2 (dashed lines) from proteome (red lines) and transcriptome (blue lines) data (B) Percentage of variation accounted for by each of the seven principal components in proteome and transcriptome data (C) Values of genes along PC-1protein plotted against PC-1mRNA. Green and blue dots represent genes with significantly large difference in expression trends (/PC-1protein−PC-1mRNA/≥2). Of these, blue dots indicate those which are likely to exhibit opposing trends (2nd and 4th quadrants). All other genes are shown as red dots.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC2367054&req=5

pone-0002097-g004: Principal component analysis of transcriptome and proteome data.(A) ‘Loadings’ (eigenvector) plot for the first two principal component axes – PC-1 (solid lines) and PC-2 (dashed lines) from proteome (red lines) and transcriptome (blue lines) data (B) Percentage of variation accounted for by each of the seven principal components in proteome and transcriptome data (C) Values of genes along PC-1protein plotted against PC-1mRNA. Green and blue dots represent genes with significantly large difference in expression trends (/PC-1protein−PC-1mRNA/≥2). Of these, blue dots indicate those which are likely to exhibit opposing trends (2nd and 4th quadrants). All other genes are shown as red dots.
Mentions: We sought to identify the major trends in both mRNA and protein dynamics using Principal Component Analysis (PCA). Of the 894 genes selected earlier, we chose to analyze only 798 gene products for which quantification ratios from at least four out of eight time points could be determined from both microarray and proteomic experiments. Missing values were estimated by linear interpolation as before. Considering the eight time-point log transformed temporal gene expression profiles as data points in an eight-dimensional space, PCA enabled us to transform coordinate axes and identify the most important dimensions in this transformed space (dimensionality reduction thereby simplifying the time-course data). Independent PCA of both mRNA and protein data from the 798 gene products indicated that the first two principal components (PC-1mRNA and PC-1protein) account for over 85% of the total variance in the either dataset (Figure 4B). The loadings plots (eigenvectors) for each of these two principal components were remarkably well correlated between mRNA and protein data (Figure 4A) implying that the major trends in both proteome and transcriptome domains were strikingly similar. Furthermore, the principal component 1 represented the tendency of genes to be either up- or down-regulated during stationary phase adaptation. This component alone accounted for over 75% of the total variance in either dataset, and therefore, appears to be the most prominent pattern in our data. The other major component, PC-2, which accounted for an additional 6–9% variance, represents the tendency of certain genes to be transiently up- or down-regulated primarily during the transition between exponential and stationary phase.

Bottom Line: Many biological processes are intrinsically dynamic, incurring profound changes at both molecular and physiological levels.Despite this overall correlation, by employing a systematic concordance analysis, we estimated that over 30% of the analyzed genes likely exhibited significantly divergent patterns, of which nearly one-third displayed even opposing trends.Our observations suggest that differences between mRNA and protein synthesis/degradation mechanisms are prominent in microbes while reaffirming the plausibility of such mechanisms acting in a concerted fashion at a protein complex or sub-pathway level.

View Article: PubMed Central - PubMed

Affiliation: Department of Chemical Engineering and Materials Science, University of Minnesota, Minneapolis, Minnesota, United States of America.

ABSTRACT
Many biological processes are intrinsically dynamic, incurring profound changes at both molecular and physiological levels. Systems analyses of such processes incorporating large-scale transcriptome or proteome profiling can be quite revealing. Although consistency between mRNA and proteins is often implicitly assumed in many studies, examples of divergent trends are frequently observed. Here, we present a comparative transcriptome and proteome analysis of growth and stationary phase adaptation in Streptomyces coelicolor, taking the time-dynamics of process into consideration. These processes are of immense interest in microbiology as they pertain to the physiological transformations eliciting biosynthesis of many naturally occurring therapeutic agents. A shotgun proteomics approach based on mass spectrometric analysis of isobaric stable isotope labeled peptides (iTRAQ) enabled identification and rapid quantification of approximately 14% of the theoretical proteome of S. coelicolor. Independent principal component analyses of this and DNA microarray-derived transcriptome data revealed that the prominent patterns in both protein and mRNA domains are surprisingly well correlated. Despite this overall correlation, by employing a systematic concordance analysis, we estimated that over 30% of the analyzed genes likely exhibited significantly divergent patterns, of which nearly one-third displayed even opposing trends. Integrating this data with biological information, we discovered that certain groups of functionally related genes exhibit mRNA-protein discordance in a similar fashion. Our observations suggest that differences between mRNA and protein synthesis/degradation mechanisms are prominent in microbes while reaffirming the plausibility of such mechanisms acting in a concerted fashion at a protein complex or sub-pathway level.

Show MeSH
Related in: MedlinePlus