Synopsis

Drug-induced mitochondrial dysfunction is of major clinical interest. An NMR-compatible bioreactor system was used to investigate the oxygen consumption of HepG2 cells by measuring T₁ of perflubron. The oxygen consumption was measured by stopping the perfusion. This resulted in a reduction of the oxygen concentration from (19.29 ± 0.96)% to (11.41 ± 0.92)% in 104.24 min. Our results might allow for detecting a drug-induced mitochondrial dysfunction of cells in a well-controlled environment by using the two compartments of the bioreactor.

Purpose:

Materials and Methods:

Experiments were performed on a 9.4 T preclinical small animal scanner (Bruker Biospec 94/20, Ettlingen, Germany) using ParaVision 6.0.1 with a linear polarized ¹H/¹⁹F Bruker surface coil. The bioreactor³ contained a 3D culture in a microcavity area (DYNARRAYS(R)) separated into two compartments which were actively perfused (400 μl/min) under normoxic conditions (Fig.1). The HepG2 cells (ATCC, HB-8065, Manassas, USA) were prepared according to^3-5.

The hydrophobicity of perflubron² required an emulsification. Therefore, a heat-sterilized emulsion was created in-house by a high shear mixing device and a sonication bath. The formulation of the emulsion was: 20% v/v perflubron (Apollo Scientific, Bredbury, UK), 2% w/v soybean lecithin (dm-drogerie markt, Karlsruhe, Germany), 2% w/v perfluorodecyl bromide (Apollo Scientific), and distilled water. The total amount of bioreactor-medium was 80 ml (15 ml emulsion and 65 ml cell culture medium).

To convert T₁ values into oxygen partial pressure (pO2), calibration measurements were performed. For this purpose, the fluid was bubbled with 0%, 10%, 21%, 60% and 100% O₂ by simultaneously adapting the N₂ concentration for two hours. The temperature dependence was evaluated for 32.5, 35.0, 37.5 and 40 °C. The measurement was performed when no temperature change had been observed for 30 min.

The perfusion shortened the T₁ value and therefore led to a wrong oxygen concentration. To remove this influence, the change of T₁ was determined by a perfusion stop with an empty chip. Then the time period to reach a constant T₁ value was used for the removal of the first T₁ values during the perfusion stop with cells (oxygen consumption measurement).

For T₁ measurements, a non-localized inversion recovery sequence and a FAIR-RARE imaging sequence (T₁ mapping) were used, followed by a 3-parameter least squares fit. For the calibration measurements, the mean T₁ values of each T₁ mapping were used for the fit of:

$$R_1 = A + B \cdot T + C \cdot pO2 + D \cdot T \cdot pO2$$

where $$$R_1 = 1/T_1$$$ is the relaxation rate, T is the temperature [°C] and A, B, C, and D are the fit parameters. Before each bioreactor experiment, T₁ of the CF₃ and CF₂Br resonance were separately measured to solve the dependence on T and pO2. For each experiment, only T₁ of the CF₃ resonance was measured.

For perflubron tracking, a SinglePulse sequence was used, followed by numerical integration of the CF₃ resonance. Table 1 lists the sequence parameters for each experiment. The analysis was done in Matlab R2015a (The MathWorks Inc, Natick, USA).

Results and Discussion:

Table 2 (literature values included^6,7) lists the calibration parameters for both resonances. The higher magnetic field changed parameter A for CF₂Br whereas the other parameters did not change much.

The inflow of the fluorinated medium led to a steep signal rise after 32 min (Fig.2). The placement of the pump and the heating bath outside the scanner room led to this late signal rise.

The influence of the perfusion on T₁ was gone after 23.07 min (Fig.3a). Therefore, the first three T₁ values during the oxygen consumption measurement were removed (Fig.3b). By replacing the empty chip with the cell chip, a reduction in the oxygen concentration from (19.29 ± 0.96)% to (11.41 ± 0.92)% in 104.24 min during a perfusion stop was observed (Fig.3b). The oxygen consumption could be completely attributed to the presence of cells as no oxygen consumption was measured without cells (Fig.3a). This novel method allowed for the determination of the oxygen consumption without the need for adapting the bioreactor setup⁸.

Conclusion:

Acknowledgements

References

1 Hussaini H, et al., Idiosyncratic drug-induced liver injury: an overview. Expert Opin Drug Saf 2007;6:673–684.

2 Riess J, Oxygen Carriers(“Blood Substitues”)Raison d’Etre, Chemistry, and Some Physiology Blut ist ein ganz besondrer Saft. Chem Rev, 2001;101:2797-2920.

3 Gottwald E, et al., Characterization of a chip-based bioreactor for three-dimensional cell cultivation via Magnetic Resonance Imaging. Z Med Phys, 2013;23:102-110.

4 Gottwald E, et al., A chip-based platform for the in vitro generation of tissues in three-dimensional organization. Lab Chip, 2007;7:777-785.

5 Altmann B, et al., The three-dimensional cultivation of the carcinoma cell line HepG2 in a perfused chip system leads to more differentiated phenotype of the cells compared to monolayer culture. Biomed Mater, 2008;3:034120.

6 Mason R, et al., A noninvasive assessment of myocardial oxygen tension: 19f nmr spectroscopy of sequestered perfluorocarbon emulsion. Magn Reson Med, 1992;27:310–317.

7 Barker B, et al., Oxygen tension mapping with F‐19 echo‐planar MR imaging of sequestered perfluorocarbon. J Magn Reson Imaging, 1994;4:595–602.

8 Pilatus U, et al., Real-time measurements of cellular oxygen consumption, pH, and energy metabolism using nuclear magnetic resonance spectroscopy. Magn Reson Med, 2001;45:749–755.