Limits...
Achieving Metabolic Flux Analysis for S. cerevisiae at a Genome-Scale: Challenges, Requirements, and Considerations.

Gopalakrishnan S, Maranas CD - Metabolites (2015)

Bottom Line: However, the necessary precautions, computational challenges, and minimum data requirements for successful analysis remain poorly established.Organism-specific challenges such as the impact of compartmentalization of metabolism, variability of biomass composition, and the cell-cycle dependence of metabolism are discussed.Identification of errors arising from incorrect gene annotation and suggested alternate routes using MFA are also highlighted.

View Article: PubMed Central - PubMed

Affiliation: Department of Chemical Engineering, The Pennsylvania State University, University Park, PA, USA. sxg375@psu.edu.

ABSTRACT
Recent advances in 13C-Metabolic flux analysis (13C-MFA) have increased its capability to accurately resolve fluxes using a genome-scale model with narrow confidence intervals without pre-judging the activity or inactivity of alternate metabolic pathways. However, the necessary precautions, computational challenges, and minimum data requirements for successful analysis remain poorly established. This review aims to establish the necessary guidelines for performing 13C-MFA at the genome-scale for a compartmentalized eukaryotic system such as yeast in terms of model and data requirements, while addressing key issues such as statistical analysis and network complexity. We describe the various approaches used to simplify the genome-scale model in the absence of sufficient experimental flux measurements, the availability and generation of reaction atom mapping information, and the experimental flux and metabolite labeling distribution measurements to ensure statistical validity of the obtained flux distribution. Organism-specific challenges such as the impact of compartmentalization of metabolism, variability of biomass composition, and the cell-cycle dependence of metabolism are discussed. Identification of errors arising from incorrect gene annotation and suggested alternate routes using MFA are also highlighted.

No MeSH data available.


Related in: MedlinePlus

Corrections for metabolite pool dilution from various sources. Rapid exchange of intracellular and extracellular CO2 introduces unlabeled carbons into the metabolic network. Although its impact decreases with time, it still plays a significant role during mid-exponential phase [88]. Dilution of compartmental amino acid pools arises from pre-existing pools. The final measured labeling distribution is the average labeling across all compartmental pools.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4588810&req=5

metabolites-05-00521-f001: Corrections for metabolite pool dilution from various sources. Rapid exchange of intracellular and extracellular CO2 introduces unlabeled carbons into the metabolic network. Although its impact decreases with time, it still plays a significant role during mid-exponential phase [88]. Dilution of compartmental amino acid pools arises from pre-existing pools. The final measured labeling distribution is the average labeling across all compartmental pools.

Mentions: Inability to completely resolve all fluxes included within the GSM model for S. cerevisiae occurs due to lack of probing techniques capable of obtaining compartment-specific labeling distribution of metabolites. Instead, existing procedures generate a pool-size-weighted average labeling distribution, which must be analyzed using corrections shown in Figure 1. This results in a degeneracy in compartment-specific labeling distributions, which is reflected in expanded flux ranges in the vicinity of that metabolite. Additional corrections to be considered during MFA of a eukaryote such as yeast include dilution by unlabeled CO2 from aeration and pre-existing metabolite pools. Furthermore, catabolism of storage compounds also adds unlabeled carbons, which may alter the estimated flux distributions if not properly accounted for. The longer doubling time of yeast results in slower biomass labeling, which in turn delays the attainment of isotopic steady-state. From the experimental perspective, the above described factors contribute to resolution loss, thereby affecting MFA flux inference. However, recent efforts have confirmed that loss of resolution can also arise from structural identifiability issues existing within the metabolic network [30]. In S. cerevisiae, loss of resolution can result from the existence of the methylglyoxal pathway and the γ-aminobutyrate pathway as alternate routes to lower glycolysis and TCA cycle, respectively, resulting in loss of resolution between these two pathways. A previous study has already attributed local flux range expansion to the presence of such alternate pathways in E. coli. In addition to this, intracellular compartmentalization of metabolism introduces metabolic cycles such as the malate shuttle [87], which are typically unresolvable using metabolite balancing techniques alone. The identification of additional metabolic loops arising from compartmentalization and their corresponding resolution criteria requires an in-depth analysis of the generated GSM mapping model of S. cerevisiae.


Achieving Metabolic Flux Analysis for S. cerevisiae at a Genome-Scale: Challenges, Requirements, and Considerations.

Gopalakrishnan S, Maranas CD - Metabolites (2015)

Corrections for metabolite pool dilution from various sources. Rapid exchange of intracellular and extracellular CO2 introduces unlabeled carbons into the metabolic network. Although its impact decreases with time, it still plays a significant role during mid-exponential phase [88]. Dilution of compartmental amino acid pools arises from pre-existing pools. The final measured labeling distribution is the average labeling across all compartmental pools.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4588810&req=5

metabolites-05-00521-f001: Corrections for metabolite pool dilution from various sources. Rapid exchange of intracellular and extracellular CO2 introduces unlabeled carbons into the metabolic network. Although its impact decreases with time, it still plays a significant role during mid-exponential phase [88]. Dilution of compartmental amino acid pools arises from pre-existing pools. The final measured labeling distribution is the average labeling across all compartmental pools.
Mentions: Inability to completely resolve all fluxes included within the GSM model for S. cerevisiae occurs due to lack of probing techniques capable of obtaining compartment-specific labeling distribution of metabolites. Instead, existing procedures generate a pool-size-weighted average labeling distribution, which must be analyzed using corrections shown in Figure 1. This results in a degeneracy in compartment-specific labeling distributions, which is reflected in expanded flux ranges in the vicinity of that metabolite. Additional corrections to be considered during MFA of a eukaryote such as yeast include dilution by unlabeled CO2 from aeration and pre-existing metabolite pools. Furthermore, catabolism of storage compounds also adds unlabeled carbons, which may alter the estimated flux distributions if not properly accounted for. The longer doubling time of yeast results in slower biomass labeling, which in turn delays the attainment of isotopic steady-state. From the experimental perspective, the above described factors contribute to resolution loss, thereby affecting MFA flux inference. However, recent efforts have confirmed that loss of resolution can also arise from structural identifiability issues existing within the metabolic network [30]. In S. cerevisiae, loss of resolution can result from the existence of the methylglyoxal pathway and the γ-aminobutyrate pathway as alternate routes to lower glycolysis and TCA cycle, respectively, resulting in loss of resolution between these two pathways. A previous study has already attributed local flux range expansion to the presence of such alternate pathways in E. coli. In addition to this, intracellular compartmentalization of metabolism introduces metabolic cycles such as the malate shuttle [87], which are typically unresolvable using metabolite balancing techniques alone. The identification of additional metabolic loops arising from compartmentalization and their corresponding resolution criteria requires an in-depth analysis of the generated GSM mapping model of S. cerevisiae.

Bottom Line: However, the necessary precautions, computational challenges, and minimum data requirements for successful analysis remain poorly established.Organism-specific challenges such as the impact of compartmentalization of metabolism, variability of biomass composition, and the cell-cycle dependence of metabolism are discussed.Identification of errors arising from incorrect gene annotation and suggested alternate routes using MFA are also highlighted.

View Article: PubMed Central - PubMed

Affiliation: Department of Chemical Engineering, The Pennsylvania State University, University Park, PA, USA. sxg375@psu.edu.

ABSTRACT
Recent advances in 13C-Metabolic flux analysis (13C-MFA) have increased its capability to accurately resolve fluxes using a genome-scale model with narrow confidence intervals without pre-judging the activity or inactivity of alternate metabolic pathways. However, the necessary precautions, computational challenges, and minimum data requirements for successful analysis remain poorly established. This review aims to establish the necessary guidelines for performing 13C-MFA at the genome-scale for a compartmentalized eukaryotic system such as yeast in terms of model and data requirements, while addressing key issues such as statistical analysis and network complexity. We describe the various approaches used to simplify the genome-scale model in the absence of sufficient experimental flux measurements, the availability and generation of reaction atom mapping information, and the experimental flux and metabolite labeling distribution measurements to ensure statistical validity of the obtained flux distribution. Organism-specific challenges such as the impact of compartmentalization of metabolism, variability of biomass composition, and the cell-cycle dependence of metabolism are discussed. Identification of errors arising from incorrect gene annotation and suggested alternate routes using MFA are also highlighted.

No MeSH data available.


Related in: MedlinePlus