Intra- and extracellular real-time analysis of perfused fibroblasts using an NMR bioreactor: A pilot study

Abstract

Introduction: Metabolomic discrimination of different mitochondrial defects is challenging. We describe an NMR-based bioreactor allowing real-time intra- and extracellular metabolic investigation of perfused fibroblasts.

Objectives: The objective of this study is (I) determining whether metabolic investigations of perfused fibroblasts overall and separated for intra- and extracellular contributions by real-time NMR allows for discrimination of different representative mitochondrial defects in a feasibility study and (II) gaining insight into physiological consequences of mitochondrial dysfunction in basal condition and during glycolysis inhibition.

Methods: Overall, intra- and extracellular metabolomes of malate dehydrogenase 2 (MDH2), pyruvate dehydrogenase (PDH), complex I (CI) deficient fibroblasts, and control fibroblasts were investigated under standard culture conditions and under glycolysis inhibition. In addition to "overall" metabolite quantification, intra- and extracellular metabolic contributions were separated based on diffusion rate differences.

Results and discussion: Overall metabolites: Chemometric analysis of the entire metabolome revealed good separation between control, PDH and MDH2, while CI was less well separated. However, mixed intra- and extracellular changes complicated interpretation of the cellular metabolism. Intra- and extracellular metabolites: Compartment specific chemometrics revealed possibly augmenting metabolomic separation between control and deficient cell lines under basal and inhibition condition. All mitochondrial defects exhibited upregulation of glycolytic metabolism compared to controls. Inhibition of glycolysis resulted in perturbations of other metabolic pathways such as glutaminolysis, alanine, arginine, glutamate, and proline metabolism. MDH2 showed upregulation of alanine and glutamate metabolism, while the CI defect revealed lower intracellular arginine and downregulation of glutamate and arginine-dependent proline synthesis.

Conclusion: Discrimination of intra- and extracellular metabolic contributions helps understanding the underlying mechanisms of mitochondrial disorders, uncovers potential metabolic biomarkers, and unravels metabolic pathway-specific adaptations in response to metabolic perturbations.

Keywords: 2‐deoxy‐glucose; NMR; bioreactor; extracellular; intracellular; mitochondrial dysfunction.

FIGURE 1

Study protocol and instrumental setting. 10 million FBs were embedded in a 3D collagen‐based matrix and perfused with media at a flow rate of 0.1 mL min⁻¹. NMR measurements were performed under both basal and inhibition conditions upon the addition of 5.55 mM 2‐DG in the bioreactor inlet. Each condition lasted approximately 3 h. From the addition of 2‐DG in the bioreactor inlet, a 1 h delay before initiating data collection allowed sufficient time for the inhibitor to reach a steady state concentration within the bioreactor tube. A sample from the inlet (control) and out‐flowing supernatants from the outlet (n = 2 in basal condition (samples a, b), n = 1 at the end of the inhibition condition (sample c)) were collected in Eppendorf tubes and stored at −80°C until the following measurement.

FIGURE 2

Chemometric analysis. (A, B) Unsupervised PCA of overall metabolites in the basal and inhibition condition, respectively. (C, D) Supervised oPLS‐DA of overall metabolites in the basal (2 LVs) and inhibition (2 LVs) condition, respectively. Ellipsoids correspond to a 95% confidence interval.

FIGURE 3

Chemometric analysis. Unsupervised principal component analysis (PCA) of intracellular metabolites separated for cell line and condition.

FIGURE 4

(A) Overall, extra‐, and intracellular glucose, pyruvate, and lactate as well as calculated extra‐ and intracellular Lac:Pyr ratio. (B) Overall, extra‐, and intracellular glutamine and glutamate as well as calculated extra‐ and intracellular glutamate to glutamine ratio. Statistical significance between groups was calculated by two‐way ANOVA followed by Benjamini and Hochberg posttest. (*) p < 0.05, (**) p < 0.01, (***) p < 0.001, (****) p < 0.0001. *(yellow): Significance for the same cell line between basal and inhibition condition; *(orange): Significance between different cell lines in basal condition; *(brown): Significance between different cell lines in inhibition condition.

FIGURE 5

Schematic representation of intracellular metabolite entrances to the TCA cycle. Statistical significance between groups was calculated by two‐way ANOVA followed by Benjamini and Hochberg posttest. (*) p < 0.05, (**) p < 0.01, (***) p < 0.001, (****) p < 0.0001. *(yellow): Significance for the same cell line between basal and inhibition condition; *(orange): Significance between different cell lines in basal condition; *(brown): Significance between different cell lines in inhibition condition.