Overview of 13c Metabolic Flux Analysis: Insights Into Cellular Pathways

Metabolic Flux Analysis (MFA) is an important tool to study the material flow in the cell metabolic network. By quantitatively analyzing the flux ( i.e., reaction rate ) of various metabolic reactions of cells under specific conditions, we can deeply analyze the metabolic activities and regulatory mechanisms of cells in different growth states. In recent years, with the development of isotope labeling technology and computational biology, 13C metabolic flux analysis ( 13C-MFA ), as a powerful tool, has been widely used in metabolic studies of bacteria, yeast and mammalian cells. By tracking the distribution of 13C-labeled substrates in the metabolic network, combining mathematical models and statistical methods, 13C-MFA can accurately quantify the flux map, which provides important theoretical basis and technical support for metabolic engineering, disease research and biological manufacturing. In this paper, the basic principles and experimental design methods of 13C-MFA will be systematically discussed, and the application of bacterial MFA will be described as an example. At the same time, the latest research progress in mammalian cell culture will be discussed in depth.

13C-metabolic flux analysis (13C-MFA)

13C-MFA is a technique to track and measure carbon source flow in intracellular metabolic pathways (MPs) by using 13C-labeled compounds (such as glucose and amino acids). 13C is a stable isotope of carbon. By adding 13C-containing substrates to the culture medium, cells absorb these labeled substrates and metabolize them into various metabolites. Through MS, tandem MS and NMR and other analytical techniques, researchers can quantify the distribution of 13C labels in these metabolites, thus reconstructing the intracellular metabolic flux. 13C-MFA has always been regarded as the gold standard for quantifying the flux of living cells in metabolic engineering. This method usually provides 13C-labeled substrates (such as [1,2–13C] glucose) to growing cells by conducting one or more tracer experiments until 13c-labeled carbon is completely mixed with metabolites and macromolecules (such as protein, RNA and glycogen) in the cells. On this basis, 13C-MFA can help to infer the carbon flow direction of different MPs in cells by analyzing the changes of these isotope labeling patterns.

Metabolic network of the upper glucose metabolism for 13C-MFA (Hogg M et al., 2023).

13C labeling pattern is highly sensitive to relative pathway flux, and different flux distributions will lead to different labeling patterns. Therefore, the metabolic flux can be inferred from the measured isotope labeling pattern. Compared with traditional MFA, the main advantage of 13C-MFA is that isotope labeling measurement provides a large number of redundant restrictions for flux estimation. For example, a typical tracer experiment may produce 50 to 100 isotope marker measurements to estimate about 10 to 20 independent metabolic fluxes. Therefore, in 13C-MFA, the number of measured data far exceeds the number of estimated flux parameters. This redundancy greatly improves the accuracy of flux estimation and enhances the confidence in the estimation results. Generally speaking, the five basic steps of 13C-MFA are: (1) Experimental design ; (2) Tracer experiment ; (3) Isotopic labeling measurement ; (4) Flux estimation ; (5) Statistical analysis .

13C metabolic flux analysis flow chart (Tian B et al., 2022).

Process of 13c Metabolic Flux Analysis

1. Tracer selection and experimental design

According to the research object, growth conditions and carbon substrate, 13C labeled substrate was selected. In the early days, single labeled substrates such as [1-13C] glucose were often used (about $100/g), but at present, double labeled substrates such as [1,2-13C] glucose are recommended (about $600/g) because they can significantly improve the accuracy of flux estimation. For microorganisms, commonly used carbon sources are glucose, acetate and glycerol, among which glucose is the most common because it is easily absorbed by many microorganisms and has rich MPs. Plants mainly rely on glucose or other carbohydrates produced by photosynthesis, while mammalian cells often use glucose, lactic acid or glutamic acid as carbon sources. MPs and growth conditions of different organisms (such as temperature, oxygen concentration, culture medium composition, etc.) will affect the utilization mode of carbon sources, so it is necessary to select suitable tracers according to research needs.

2. Steady-state culture and sample collection

Ensuring metabolic and isotopic homeostasis is the key to the success of the experiment. To ensure the stability of the metabolic system, the following three methods are usually adopted:

• Prolonging the incubation time at constant temperature: Usually, the incubation time is more than 5 residence times to ensure the system to reach the stable state and realize the isotope steady state. In this process, cells continue to grow and consume carbon sources, but the flux of metabolic pathway remains constant, thus providing a reliable basis for subsequent analysis.
• Batch culture experiment: By keeping the cell growth rate constant (for example, in exponential growth period), the metabolic flux can be stabilized. This method enables cells to continue their metabolic activities under controllable growth conditions, and ensures the repeatability and accuracy of experimental data.
• Optimizing the number of tracer experiments: The number of tracer experiments depends on the required flux resolution. Increasing the number of experiments and labeling measurements can significantly improve the accuracy of flux estimation. According to the research of Crown et al, the uncertainty of flux estimation can be controlled within 5% by two parallel labeling experiments, which meets the accuracy requirements of most studies.

3. Isotopic marker measurements

The third step of 13C-MFA is to measure and collect isotope labeling data of samples, which is the key link to accurately quantify metabolic flux. At present, GC-MS is the most commonly used analytical method, which can determine the isotope distribution of metabolites with high precision and provide reliable data support for 13C-MFA. In addition, the flux resolution can be further improved by combining various analysis techniques, mainly including the following methods:

• NMR: As a powerful analytical tool, NMR can provide detailed information of metabolite structure and isotope labeling. Although its resolution is usually lower than that of GC-MS, NMR has unique advantages in the analysis of liquid samples and can provide global metabolic information.
• LC-MS/MS: This technique is excellent in liquid sample analysis, which can not only analyze complex metabolite spectrum, but also significantly improve the resolution of sample separation. LC-MS/MS combines the high separation ability of liquid chromatography with the qualitative and quantitative advantages of ms, which makes the analysis results more accurate and reliable.
• GC-MS/MS: Through multiple ms analysis, tandem ms technology significantly improved the detection sensitivity and resolution. GC-MS/MS is especially suitable for complex metabolic network research and other application scenarios that require high accuracy.

4. Flux estimation and model solution

The core of this step is to deduce the metabolic flux parameters through nonlinear regression, so as to best fit the isotope labeling pattern and external rate data measured by experiments. Because metabolic flux is influenced by many factors, such as metabolite concentration, pathway branching, isotope labeling mode and reaction rate, the nonlinear regression problem is highly complex. Therefore, researchers have developed computational tools (such as Metran, INCA, OpenFLUX2, etc.) based on the framework of EMU(Elementary Metabolic Units), and the process of flux estimation is significantly simplified by decomposing the complex metabolic network into basic units for modular analysis. The standardized method of EMU framework makes the modeling and solution of metabolic network more operable and repeatable.

5. Statistical analysis and verification

In order to ensure the reliability of flux estimation, it is necessary to verify the fitting effect of the model by statistical methods:

(1) Residual sum of squares (SSR) evaluation: SSR reflects the deviation between model prediction and experimental data, and the smaller the value, the higher the fitting degree. The minimized SSR should obey χ distribution (degree of freedom: number of data points n-number of parameters p). By setting the confidence level (such as α=0.05), the reasonable range of SSR is determined (χ α/2 (n-p) ≤ SSR ≤ χ 1-α/2 (n-p)). If SSR is beyond this range, the following problems need to be investigated:

• Metabolic model is incomplete or reaction reversibility is not set correctly.
• Measurement error or signal noise interference
• The quality of isotope labeling data is insufficient.

(2) Confidence interval calculation: Quantify the uncertainty of flux estimation by the following methods:

• Sensitivity analysis: evaluate the influence of small changes in flux parameters on SSR and determine the sensitivity of key flux.
• Monte Carlo simulation: based on the distribution of flux solutions generated by random sampling, the confidence interval is calculated statistically to provide probabilistic reliability evaluation of the results.

If the SSR test fails, it is necessary to return to the fourth step to readjust the model or experimental design until the statistical standard is met. This closed-loop process ensures the scientific rigor of 13C-MFA results.

Examples of bacterial metabolic flux analysis

13C-MFA is widely used in bacterial metabolism research, especially in metabolic engineering of industrial fermentation and microbial production.

• Extremely thermophilic bacteria refer to those microorganisms that grow best at 75°C or higher, and they can usually survive in extreme environments, which makes them an important object in the fields of metabolic mechanism, enzymology, industrial application and so on. Extremely thermophilic bacteria can be used to treat pollutants and wastes such as heavy metals or organic pollutants in the environment; Is the source of thermostable enzymes; Thermophilic bacteria can use carbon sources extracted from renewable raw materials for biotransformation to produce biofuels (such as ethanol) and various chemicals. In this study, 13C-MFA was used to verify the metabolic networks of three different kinds of extremely thermophilic bacteria Geobacillus sp. LC300, Thermophilus HB8 and Rhodothermus marinus DSM 4252. Through six parallel labeling experiments, using 13C- glucose tracers labeled at different positions, combined with the data from amino acid isotope labeling, it was found that the main MPs of Geobacillus sp. LC300 were glycolysis, oxPPP and TCA cycle. The glycolysis and TCA cycle of T. thermophilus are the main active pathways, while other pathways are basically inactive. Through 13C-MFA, it was found that thermophilic carboxylase and phosphoenolpyruvate carboxykinase were active simultaneously in this organism. R. marinus is similar to T. thermophilus, and glycolysis and TCA cycle are the main pathways, while other pathways are almost inactive. The synthetic routes of Lys, Ser and Gly of Geobacillus sp. LC300, T. thermophilus and R. marinus are basically the same, and Lys is synthesized by the classical DAP route, Ser comes from 3PG, and Gly is synthesized by Ser and Thr. However, the synthetic pathway of isoleucine (Ile) is different. There are also differences in carbon balance and MPs. (1) Carbon balance: most of glucose is converted into biomass (47-60%), and the rest is converted into CO2(28-49%) and acetate (12% for Geobacillus LC300). (2)NADH metabolism: Geobacillus sp. LC300 produces NADH through glycolysis and TCA cycle, while T. thermophilus and R. marinus mainly rely on TCA cycle to produce NADH. (3)NADPH metabolism: Geobacillus sp. LC300 produces NADPH by oxidizing pentose phosphate pathway (oxPPP), T. thermophilus by TCA cycle, and R. marinus by transaminase. (4)ATP metabolism: Three thermophilic bacteria mainly produce ATP through oxidative phosphorylation, and glycolysis also contributes (Cordova LT et al., 2017).
• Aerobic culture experiment of Escherichia coli in micro bioreactor to explore the effects of different isotope tracers on 13C-MFA. Eight tracers are used, including four common glucose tracers (such as [1,2-13C] glucose and [U-13C] glucose) and four new tracers (such as the mixture of [2,3-13C] glucose and [4,5,6-13C] glucose). During the exponential growth period, the collected biomass samples were used for isotope analysis, and the mass isotope distribution of 14 amino acids was determined by GC-MS technology. A total of 1246 mass isotope abundances were measured to provide data for the subsequent 13C-MFA. In this paper, the combination 13C-MFA (that is, COMPLETE-MFA) is carried out by combining 14 parallel experimental data, which is the first time that so many isotope labeling experiments have been successfully integrated into 13C-MFA. The analysis results show that the model fits all the data effectively within a statistically acceptable range, and accurate flux estimation and 95% confidence interval are obtained. Different isotope tracers were used to test the effect of COMPLETE-MFA on metabolic flux of Escherichia coli. The results show that different tracers have different effects on the accuracy of metabolic flux. For example, [4,5,6-13C] glucose has the best accuracy for TCA circulating flux, while [2,3-13C] glucose is the only tracer that can successfully estimate all 10 net free fluxes in the model. Commonly used isotope tracers, such as the 4: 1 mixture of [1-13C] glucose and [U-13C] glucose, perform best, especially in the upper flux of central metabolism (such as glycolysis, pentose phosphate pathway and Entner-Doudoroff pathway), but the estimation of TCA circulation flux and anaerobic reaction is poor, and among the 14 tracers, no single tracer performs best in all metabolic fluxes. The best single tracer is [5-13C] glucose, which shows good accuracy in many key fluxes and can determine more exchange fluxes (Crown SB et al., 2015).
• Phenol is an important chemical, which is usually produced by benzene in fossil resources. Tyrosine phenol lyase (TPL) in some microorganisms catalyzes the formation of phenol. At present, we are exploring the production of phenol from renewable resources through microbial fermentation. However, the tolerance of E.coli to phenol is weak, especially at higher concentrations, and its growth will be limited. 13C-MFA was used to compare the carbon flux distribution of E.coli BW25113 under different phenol concentrations, and the influence of phenol on central carbon metabolism was revealed (under the conditions of phenol concentrations of 0%, 0.1% and 0.15%). With the increase of phenol concentration, the specific growth rate of Escherichia coli decreased significantly. For example, at 0.1% phenol, the specific growth rate decreased by 40%, while at 0.16% phenol, it decreased by 84%. The maximum OD660 value decreased by 16% under 0.1% phenol and by 92% under 0.16% phenol, which indicated that phenol had a strong inhibitory effect on cell growth. In media containing different concentrations of phenol (0, 0.1, 0.15%), the addition of phenol has little effect on the flux of upstream pathways such as glycolysis, pentose phosphate pathway and ED pathway, but the glucose consumption and acetate production of Escherichia coli are affected. With the increase of phenol concentration, the consumption rate of glucose slows down, while the concentration of acetate increases significantly when the phenol concentration is high. A mutant strain lacking phosphotransacetylase (Δpta) was used in this study. The results showed that the growth of Δpta strain was similar to that of wild type without phenol, but it could not grow under 0.15% phenol. This shows that the overflow of acetate may be an important factor for the growth of Escherichia coli under phenol toxicity. The specific activity of citrate synthase did not change significantly under the condition of phenol addition, which indicated that the inhibitory effect of phenol might not be directly through the denaturation of the enzyme. Although the activity of citrate synthase changed little, phenol still slowed down the cell growth by affecting metabolic flux (Kitamura S et al., 2019).

For more information about the role of MFA in bacterial metabolism, please refer to "Metabolic Flux analysis of Bacteria".

MFA in mammalian cell culture

Different from bacteria, the metabolic pathway of mammalian cells is more complex and regulated by many factors, such as oxygen concentration, culture medium composition, cell proliferation state and so on. The application of 13C-MFA in mammalian cell culture can help researchers to deeply understand the energy metabolism, amino acid metabolism and adaptation to the external environment of cells.

Characteristics of mammalian cell metabolism

Mammalian cells usually face the following metabolic challenges in the process of culture:

• Aerobic and anaerobic metabolism: when oxygen supply is sufficient, mammalian cells will generate ATP through aerobic metabolism. However, under hypoxic conditions, cells will rely on anaerobic metabolism (such as lactic acid fermentation) to maintain energy supply.
• Amino acid metabolism: mammalian cells need exogenous essential amino acids, but they also use the amino acids in the culture medium for anabolism.
• Fatty acid and sugar metabolism: cells can adjust the intake and utilization of fatty acids and sugar according to environmental changes.

Application

• In mammalian cells, the commonly used 13C tracers include 13C- glucose, 13C- glutamine, 13C- propionic acid and 13C- glycerol, etc. These tracers have different MPs and labeling patterns, and are suitable for different research objectives. Common 13C tracers include 13C- glucose and 13C- glutamine, and it is very important to select the best tracer for different MPs to improve the analysis accuracy. By applying the EMU(Exchangeable Metabolic Units) method, researchers can select suitable tracers to estimate specific metabolic fluxes, such as pentose phosphate pathway (oxPPP) and pyruvate carboxylase (PC). The model considers labeled carbon sources (13C-labeled glucose and glutamine) and their roles in lactic acid synthesis. The model has two degrees of freedom, namely pentose phosphate oxide flux (oxPPP) and pyruvate carboxylase flux (PC), and the mass isotope distribution (MID) of lactic acid provides additional constraints for determining these two degrees of freedom. Each EMU basis vector represents a possible approach and quantifies its contribution to lactate labeling. The final model gives 156 basis vectors, and quantitatively analyzes how these vectors play a role in lactic acid production in different cell types (such as WT cells and PC expression cells). WT cells mainly produce lactic acid through glycolysis and oxPPP, and the contribution of glycolysis is dominant (about 93%). In PC-expressing cells, the contribution of glycolysis and oxPPP to lactic acid decreased due to the increase of malic enzyme flux, and they were replaced by anaerobic reactions (such as PC and glutamine decomposition), which accounted for a large part of the total lactic acid production. In the production of lactic acid, glutamine decomposition provides a part of the contribution, especially the characteristic basis vectors such as Gln234 and Gln345, which contribute about 3% to the production of lactic acid (Crown SB et al., 2012).
• Pulmonary fibrosis (PF) is a progressive and chronic lung disease, characterized by lung tissue damage caused by excessive collagen deposition, which eventually leads to respiratory failure. Studies show that glutamine decomposition plays an important role in PF. Glutamine is decomposed into glutamic acid by transglutaminase (GLS1), and further converted into α-ketoglutaric acid, which supplements TCA cycle. This process supports cell energy production and biosynthesis. The up-regulation of GLS1 plays a key role in PF. Tanshinone IIA is a natural compound extracted from salvia miltiorrhiza root, which has antioxidant and anti-fibrosis effects. The results showed that glutamine (2 mM and 4 mM) significantly promoted the proliferation of NIH-3T3 cells induced by TGF-β1(10 ng/mL) and increased the percentage of EdU positive cells. 13C-MFA data showed that Tan IIA regulated TCA cycle and glutamine metabolism, which led to significant changes in metabolite abundance induced by TGF-β1. Especially in the experiment of [U-13C5] glutamine labeling, Tan IIA (10 μM) can significantly reduce the flux of glutamine into TCA cycle through glutamine decomposition pathway, and inhibit the expression of key enzymes related to proline synthesis, such as GLS1, PSAT1, P5CS and PYCR1. In addition, the effect of Tan IIA is similar to CB-839, which indicates that TanIIA may interfere with the production of glutamic acid and α-KG during fibrosis by inhibiting glutamine decomposition and related MPs, and then affect the synthesis of collagen. Molecular docking analysis further showed that Tan IIA has strong binding affinity with these key enzymes (such as GLS1, PSAT1, P5CS and PYCR1), and its biological effect may be achieved by directly inhibiting the activities of these enzymes (Shan B et al., 2024).

If you want to list more functions of MFA in mammalian cell culture, please refer to "Metabolic Flux Analysis in Mammalian Cell Culture".

References

1. Cordova LT, Cipolla RM, Swarup A, Long CP, Antoniewicz MR. "13C metabolic flux analysis of three divergent extremely thermophilic bacteria: Geobacillus sp. LC300, Thermus thermophilus HB8, and Rhodothermus marinus DSM 4252." Metab Eng. 2017;44:182-190. doi: 10.1016/j.ymben.2017.10.007
2. Crown SB, Long CP, Antoniewicz MR. "Integrated 13C-metabolic flux analysis of 14 parallel labeling experiments in Escherichia coli." Metab Eng. 2015;28:151-158. doi: 10.1016/j.ymben.2015.01.001
3. Kitamura S, Toya Y, Shimizu H. "13C-Metabolic Flux Analysis Reveals Effect of Phenol on Central Carbon Metabolism in Escherichia coli." Front Microbiol. 2019;10:1010. doi: 10.3389/fmicb.2019.01010
4. Crown SB, Ahn WS, Antoniewicz MR. "Rational design of ¹³C-labeling experiments for metabolic flux analysis in mammalian cells." BMC Syst Biol. 2012 ;6:43. doi: 10.1186/1752-0509-6-43
5. Shan B, Guo C, Zhou H, Chen J. "Tanshinone IIA alleviates pulmonary fibrosis by modulating glutamine metabolic reprogramming based on [U-13C5]-glutamine metabolic flux analysis." J Adv Res. 2024 May 1:S2090-1232(24)00172-3. doi: 10.1016/j.jare.2024.04.029