Background

Mitochondria are the cell’s powerhouse and important signaling organelles. Dysfunctions either originate from primary genetic defects or are induced by signaling pathways, toxins or drugs. Thus, different mitopathogenic mechanisms have been described across all medical disciplines, highlighting the interdependence of bioenergetic function with a multitude of metabolic and signaling pathways. To our knowledge, there is no method enabling simultaneous analysis of both, mitochondrial function and metabolomic changes in living cells. Recently Bruker released a flow tube (InsightCell™), which allows for investigations of living cells with no need for special equipment using a standard 5mm NMR tube. Previously we have demonstrated the feasibility of real-time metabolic monitoring in 3D culture in a bioreactor¹. Additionally it is well known that the oxygen tension can be determined via T₁ relaxation time measurements of ¹⁹F compounds^2,3.
In this study we evaluate the effect of different perfusion rate on metabolic profiles. We aim at simultaneous detection of substrate conversion using ¹H NMR and mitochondrial respiration using ¹⁹F NMR relaxation time measurements of perfluorocarbons in a human fibroblast 3D cell culture system. A standard glycolytic stress test is finally used to test the feasibility to simultaneously cellular metabolism and mitochondrial respiration.

Methods

NMR experiments were performed on a 500MHz Bruker Avance II spectrometer. HPLC pumping system (Agilent) was used for constant perfusion (0.025 – 0.3ml/min) of substrate via the perfusion apparatus InsightCell™. For experiments with chemical inhibitors, serum free culture medium was used. In order to maintain the pH and supply of sufficient oxygen, a gas mix of 45% O₂, 50% N₂ and 5% CO₂ was applied to the bioreactor using a self-constructed gas blending system regulated by DigiFlow™ digital flowmeters (Draeger Medical). Online metabolic profiling by ¹H NMR was performed using 1D project spectra of 32 transients (≈4min) per spectrum (TE=80ms), allowing kinetic tracking of 30 intra- and extracellular metabolites under standard cell culture condition (Figure 1)⁴. ¹⁹F T₁-relaxation time measurements of perfluorotributylamine (PFTBA) were performed for oxygen quantification. Measurements were performed using a pseudo 2D inversion recovery 180°-τ-90° pulse sequence (t1pir) using 8 incremented delay times (0.005 s, 0.646 s, 1.317 s, 2.030 s, 2.804 s, 3.676 s, 4.732 s, 6.277 s) with 4 scans per time increment (total of 5min). Cell viability was controlled by trypan blue staining and cellular LDH release.

Results & Discussion

We implemented a perfused bioreactor system within the NMR spectrometer shown in figure 2 (shown before at ISMRM 2019), evaluated embedding methods for high cell density in collagen based matrix (20 million cells), viability (>90%), stability (up to 12h) and reproducibility of metabolic and oxymetric responses.

Online metabolic profiling: The level of glycolytic metabolites, glutamine as well as the level of free fatty acids were monitored during 6 hours of perfusion of 20 million fibroblasts. Six times repeated seven-minute interruption of the perfusion rate resulted in reproducible findings with significant decrease of glucose and pyruvate level and simultaneous increase of lactate levels (Figure 3B). In contrast, glutamine and free fatty acid levels did not show obvious fast periodic adaptations to short term flow changes. However also glutamine and free fatty acid levels were perfusion rate depending as shown in figure 3C. While glycolytic metabolites show immediate changes, glutamine and fatty acid compounds showed significant changes after changing the flow rate for ~90 minutes (Figure 3C).

Oxygen detection using ¹⁹F of non-emulsified PFTBA: Emulsifiers for PFTBA have been used in other bioreactor setups for convenient incorporation into the system for T₁ time determination⁵. We avoid the use of emulgators due to interference signals in the proton spectrum and possible physiological effects on embedded cells. However, initial tests demonstrated strong T₁ adaption delays of PFTBA upon O₂ changes, which appeared to depend on the size of the organic PFTBA phase. Therefore, to achieve high contact area between both liquid phases without chemical emulsifiers, we dipped a 0.27mm cotton thread for 1 hour into PFTBA to cover its surface with PFTBA. Embedding this PFTBA coated thread into the 3D cell culture scaffold proved to be a sensitive oxygen sensor responding to changes in T₁ within minutes (Figure 4).

Glycolytic stress test: In order to probe the bioreactor system sensitivity upon challenges, a glycolytic stress test was performed (Figure 5). Addition of the respiratory chain inhibitors oligomycine & rotenone lead to an increase of the determined oxygen level up to the oxygen amount of inserted medium. Simultaneous upregulation of the lactate production indicates glycolytic capacity. Subsequent addition of 2-deoxy-glucose inhibited lactate production completely. Detection of glycolytic rate and maximum glycolytic capacity provides important information on cell status and phenotype. It allows quantitative analysis of energy production of glycolysis and oxidative phosphorylation.

Conclusion

In this study we show the feasibility to simultaneously measure respirometric and metabolic data using standard 5mm NMR tube. We describe the effect of the flow rate on the metabolic activity. Importantly we show the required sensitivity to detect substrate degradation rates of major mitochondrial fuel pathways and ability to measure rapid O₂ and lactate changes as surrogate marker of oxidative phosphorylation and anaerobic glycolysis.