Limits...
Spatio-temporal dynamics of yeast mitochondrial biogenesis: transcriptional and post-transcriptional mRNA oscillatory modules.

Lelandais G, Saint-Georges Y, Geneix C, Al-Shikhley L, Dujardin G, Jacq C - PLoS Comput. Biol. (2009)

Bottom Line: This last spatio-temporal module concerns mostly mRNAs coding for basic elements of mitochondrial construction: assembly and regulatory factors.Prediction that unknown genes from this module code for important elements of mitochondrial biogenesis is supported by experimental evidence.More generally, these observations underscore the importance of post-transcriptional processes in mitochondrial biogenesis, highlighting close connections between nuclear transcription and cytoplasmic site-specific translation.

View Article: PubMed Central - PubMed

Affiliation: Dynamique des Structures et Interactions des Macromolécules Biologiques (DSIMB), INSERM UMR-S 665, Université Paris Diderot, Paris, France. gaelle.lelandais@univ-paris-diderot.fr

ABSTRACT
Examples of metabolic rhythms have recently emerged from studies of budding yeast. High density microarray analyses have produced a remarkably detailed picture of cycling gene expression that could be clustered according to metabolic functions. We developed a model-based approach for the decomposition of expression to analyze these data and to identify functional modules which, expressed sequentially and periodically, contribute to the complex and intricate mitochondrial architecture. This approach revealed that mitochondrial spatio-temporal modules are expressed during periodic spikes and specific cellular localizations, which cover the entire oscillatory period. For instance, assembly factors (32 genes) and translation regulators (47 genes) are expressed earlier than the components of the amino-acid synthesis pathways (31 genes). In addition, we could correlate the expression modules identified with particular post-transcriptional properties. Thus, mRNAs of modules expressed "early" are mostly translated in the vicinity of mitochondria under the control of the Puf3p mRNA-binding protein. This last spatio-temporal module concerns mostly mRNAs coding for basic elements of mitochondrial construction: assembly and regulatory factors. Prediction that unknown genes from this module code for important elements of mitochondrial biogenesis is supported by experimental evidence. More generally, these observations underscore the importance of post-transcriptional processes in mitochondrial biogenesis, highlighting close connections between nuclear transcription and cytoplasmic site-specific translation.

Show MeSH
Description of the EDPM algorithm.A: Description of the vectors (D,M and W) and thematrix (P) used by EDPM. The algorithm calculatesthe vector W of ω-values,using an optimization procedure (see main text and FigureS1). B: Representation of the 15 model patterns used inthis study. These models are periodic functions covering threeconsecutive cycles. The color code reflects the metabolic phasesduring which model patterns are maximal(R/B = green;R/C = blue andOx = red). C: Illustration of EDPMresults analyzing R/B, R/C and Ox sentinel genes defined in [3]. Initial vectors from the microarraydata D are plotted in orange; theM vector, obtained by multiplying ofW and P, is plotted in red. Theω-values (also referred to asω-footprints in the main text) arerepresented as barplots and model patterns are indicated with acolor code.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2690403&req=5

pcbi-1000409-g001: Description of the EDPM algorithm.A: Description of the vectors (D,M and W) and thematrix (P) used by EDPM. The algorithm calculatesthe vector W of ω-values,using an optimization procedure (see main text and FigureS1). B: Representation of the 15 model patterns used inthis study. These models are periodic functions covering threeconsecutive cycles. The color code reflects the metabolic phasesduring which model patterns are maximal(R/B = green;R/C = blue andOx = red). C: Illustration of EDPMresults analyzing R/B, R/C and Ox sentinel genes defined in [3]. Initial vectors from the microarraydata D are plotted in orange; theM vector, obtained by multiplying ofW and P, is plotted in red. Theω-values (also referred to asω-footprints in the main text) arerepresented as barplots and model patterns are indicated with acolor code.

Mentions: Our aim was to investigate in more detail the published gene expressionprofiles obtained from yeast cell cultures displaying highly periodic cyclesin the form of respiratory bursts. Starting from previous work [6], we developed the EDPM algorithm (ExpressionDecomposition based on Periodic Models) to analyze precisely gene expressionpatterns during yeast metabolic cycles. Overview of the EDPM procedure ispresented in Figure S1. The main idea is to decompose each gene expressionprofile obtained with microarray technology (vector D, Figure 1A),into a mixture of pre-defined model patterns (matrix P, Figure 1A). Forthat, the algorithm calculates a vector W of ω-values, such that the standardmultiplication of vector W and matrix P, forms a vector M that reproduces the initial vector of expression measurements D (eq. 1 below). The ω-values are determine using an optimization procedure (see below and Figure 2 for anillustration) and therefore indicate the contribution of each model patternto the expression pattern observed for a particular gene. The data vector M matches the vector D exactly if P is a perfect model of the biological:(1)


Spatio-temporal dynamics of yeast mitochondrial biogenesis: transcriptional and post-transcriptional mRNA oscillatory modules.

Lelandais G, Saint-Georges Y, Geneix C, Al-Shikhley L, Dujardin G, Jacq C - PLoS Comput. Biol. (2009)

Description of the EDPM algorithm.A: Description of the vectors (D,M and W) and thematrix (P) used by EDPM. The algorithm calculatesthe vector W of ω-values,using an optimization procedure (see main text and FigureS1). B: Representation of the 15 model patterns used inthis study. These models are periodic functions covering threeconsecutive cycles. The color code reflects the metabolic phasesduring which model patterns are maximal(R/B = green;R/C = blue andOx = red). C: Illustration of EDPMresults analyzing R/B, R/C and Ox sentinel genes defined in [3]. Initial vectors from the microarraydata D are plotted in orange; theM vector, obtained by multiplying ofW and P, is plotted in red. Theω-values (also referred to asω-footprints in the main text) arerepresented as barplots and model patterns are indicated with acolor code.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2690403&req=5

pcbi-1000409-g001: Description of the EDPM algorithm.A: Description of the vectors (D,M and W) and thematrix (P) used by EDPM. The algorithm calculatesthe vector W of ω-values,using an optimization procedure (see main text and FigureS1). B: Representation of the 15 model patterns used inthis study. These models are periodic functions covering threeconsecutive cycles. The color code reflects the metabolic phasesduring which model patterns are maximal(R/B = green;R/C = blue andOx = red). C: Illustration of EDPMresults analyzing R/B, R/C and Ox sentinel genes defined in [3]. Initial vectors from the microarraydata D are plotted in orange; theM vector, obtained by multiplying ofW and P, is plotted in red. Theω-values (also referred to asω-footprints in the main text) arerepresented as barplots and model patterns are indicated with acolor code.
Mentions: Our aim was to investigate in more detail the published gene expressionprofiles obtained from yeast cell cultures displaying highly periodic cyclesin the form of respiratory bursts. Starting from previous work [6], we developed the EDPM algorithm (ExpressionDecomposition based on Periodic Models) to analyze precisely gene expressionpatterns during yeast metabolic cycles. Overview of the EDPM procedure ispresented in Figure S1. The main idea is to decompose each gene expressionprofile obtained with microarray technology (vector D, Figure 1A),into a mixture of pre-defined model patterns (matrix P, Figure 1A). Forthat, the algorithm calculates a vector W of ω-values, such that the standardmultiplication of vector W and matrix P, forms a vector M that reproduces the initial vector of expression measurements D (eq. 1 below). The ω-values are determine using an optimization procedure (see below and Figure 2 for anillustration) and therefore indicate the contribution of each model patternto the expression pattern observed for a particular gene. The data vector M matches the vector D exactly if P is a perfect model of the biological:(1)

Bottom Line: This last spatio-temporal module concerns mostly mRNAs coding for basic elements of mitochondrial construction: assembly and regulatory factors.Prediction that unknown genes from this module code for important elements of mitochondrial biogenesis is supported by experimental evidence.More generally, these observations underscore the importance of post-transcriptional processes in mitochondrial biogenesis, highlighting close connections between nuclear transcription and cytoplasmic site-specific translation.

View Article: PubMed Central - PubMed

Affiliation: Dynamique des Structures et Interactions des Macromolécules Biologiques (DSIMB), INSERM UMR-S 665, Université Paris Diderot, Paris, France. gaelle.lelandais@univ-paris-diderot.fr

ABSTRACT
Examples of metabolic rhythms have recently emerged from studies of budding yeast. High density microarray analyses have produced a remarkably detailed picture of cycling gene expression that could be clustered according to metabolic functions. We developed a model-based approach for the decomposition of expression to analyze these data and to identify functional modules which, expressed sequentially and periodically, contribute to the complex and intricate mitochondrial architecture. This approach revealed that mitochondrial spatio-temporal modules are expressed during periodic spikes and specific cellular localizations, which cover the entire oscillatory period. For instance, assembly factors (32 genes) and translation regulators (47 genes) are expressed earlier than the components of the amino-acid synthesis pathways (31 genes). In addition, we could correlate the expression modules identified with particular post-transcriptional properties. Thus, mRNAs of modules expressed "early" are mostly translated in the vicinity of mitochondria under the control of the Puf3p mRNA-binding protein. This last spatio-temporal module concerns mostly mRNAs coding for basic elements of mitochondrial construction: assembly and regulatory factors. Prediction that unknown genes from this module code for important elements of mitochondrial biogenesis is supported by experimental evidence. More generally, these observations underscore the importance of post-transcriptional processes in mitochondrial biogenesis, highlighting close connections between nuclear transcription and cytoplasmic site-specific translation.

Show MeSH