Author Archives: sepideh2013

Final project: Manon Morin, Sepideh Dolatshahi – Integrative model of post transcriptional regulation of the metabolic transition in E.coli

The ability of the bacteria to switch from one carbon substrate to another (i.e to reorganize its metabolism) is one of the complex and highly regulated mechanisms that bacteria have developed for responding to their environment changes. The CSR (Carbon Storage Regulator) is a post-transcriptional regulator involved in the regulation of such mechanisms, the protein CsrA (main component of the CSR system) can activate or inhibit the translation of target mRNA’s and has been shown to activate the glycolytic metabolism while inhibiting the gluconeogenetic pathway. We are interested in understanding the role of the CSR system during the switch from glycolysis to gluconeogenesis.

To achieve this goal, we need to integrate the dynamical data in different levels of transcriptomics, proteomics, and metabolomics into one consistent model which could explain the interplay of the CSR system and the transition dynamics of interest. The data at hand includes time courses of interacellular metabolite concentrations measured with mass spectrometry, extracellular concentrations of glucose and acetate measured with HPLC, and enzymatic assays. And finally time series data for mRNA (transcriptomics). So far we only have data for delta-csrA.

  We understand that this modeling project is complex and needs refining and going back and forth between experiments and modeling simulations. To put this into a manageable modeling effort in the context of this course, two different approaches were taken. First, we started at the metabolic level, and only used the affinity constants (Km) from the literature and estimated the missing Km’s and all Vmax’s by fitting to the metabolomics data.  We got a relatively good fit. The next step would be to use the enzymatic assays to update the Vmax’s since we assume they scale linearly with the amount of enzymes and check to see if the predictions agree with the concentration data for the knock out strains. We expect to see disagreements which will help us learn more about the dynamics and update the structure of the model. 

 As a second strategy we looked at each enzymatic reaction one at a time, and incorporated the three levels of gene to mRNA to protein/enzyme which regulates that reaction in a sub-model. This model is supposed to help us locate the point of csrA regulation and confirm and add to what is in the literature about where csrA regulates the glycolytic and gluconeogenic pathways.

The simulations and optimizations were done in MATLAB for the first part of the project. We used COPASI for the modeling of the second part in order to explore that software and its capabilities. 

 Once Manon obtains the complete data sets, we could proceed with integrating all the sub models with the metabolic model and learn about the complex interactions of the different moving pieces of this multi-level integrative model. 

 

 

Advertisements

Project Idea: Post transcriptional regulation in metabolic transition in E.coli, Manon Morin, Sepideh Dolatshahi [Albert]

Question: What is the role of the post transcriptional regulation by the CSR system during the glucose/acetate switch?

We are interested in specifics of how csrA regulates both Glycolysis and Gluconeogenesis and the transition between those two states. (Literature has some examples of specific inhibition targets, we want a more complete dynamic picture)

Data: 

  • Enzymatic assays which show us how enzyme activities change from WT to knockout (proteomics)
  • Time series of metabolite concentrations (metabolomics)
  • Time series of gene expression data (genomics)

Modeling approach: 

We start with a manageable piece of the whole picture which spreads along metabolomics, proteomics, and transcriptomics levels.

After long discussions, we chose to initially focus on one reaction (enzyme) at a time, and build a model incorporating the three levels of regulation.

So far we have data for the wild type strain. We are planning to do parameter estimation to fit the related experimental data. And the behavior in the  strain could be predicted by removing the CsrA protein from the model. The predictive results could be tested using the data from  (will be collected in Sep-Oct).

Image

Sepideh Dolatshahi, Georgia Institute of Technology

Sepideh