Continous, non-invasive monitoring of oxygen consumption in a parallelized microfluidic in vitro system provides novel insight into the response to nutrients and drugs of primary human hepatocytes

Abstract

Oxygen plays a fundamental role in cellular energy metabolism, differentiation and cell biology in general. Consequently, in vitro oxygen sensing can be used to assess cell vitality and detect specific mechanisms of toxicity. In 2D in vitro models currently used, the oxygen supply provided by diffusion is generally too low, especially for cells having a high oxygen demand. In organ-on-chip systems, a more physiologic oxygen supply can be generated by establishing unidirectional perfusion. We established oxygen sensors in an easy-to-use and parallelized organ-on-chip system. We demonstrated the applicability of this system by analyzing the influence of fructose (40 mM, 80 mM), ammonium chloride (100 mM) and Na-diclofenac (50 µM, 150 µM, 450 µM, 1500 µM) on primary human hepatocytes (PHH). Fructose treatment for two hours showed an immediate drop of oxygen consumption (OC) with subsequent increase to nearly initial levels. Treatment with 80 mM glucose, 20 mM lactate or 20 mM glycerol did not result in any changes in OC which demonstrates a specific effect of fructose. Application of ammonium chloride for two hours did not show any immediate effects on OC, but qualitatively changed the cellular response to FCCP treatment. Na-diclofenac treatment for 24 hours led to a decrease of the maximal respiration and reserve capacity. We also demonstrated the stability of our system by repeatedly treating cells with 40 mM fructose, which led to similar cell responses on the same day as well as on subsequent days. In conclusion, our system enables in depth analysis of cellular respiration after substrate treatment in an unidirectional perfused organ-on-chip system.

Keywords: in vitro; liver; metabolism; model; organ-on-chip; oxygen; perfusion; sensors; toxicity.

Table 1. Substances used for treating cells cultured in the HepaChip-MP

Table 2. Comparison of features of the HepaChip-MP to other available systems to measure oxygen consumption in cell culture systems

Figure 1. HepaChip-MP with oxygen sensor spots and readout system. A) Schematic overview of the HepaChip-MP with 24 independent cell culture chambers in wellplate format. B) Individual cell culture chamber of the HepaChip-MP showing electrodes (yellow), capillary stop valves to control bubble free filling and three assembly ridges to culture sinusoid-like micro-tissues. Prior to cell assembly the assembly ridges are functionalized by collagen I. C) Microscopic image of cell culture chamber showing the location of the oxygen sensor spots. By measuring oxygen concentration at both the inlet and outlet of the chamber, the oxygen consumption of the microtissue can be determined. D) Photo of the readout system: optical fibers illuminating and probing the oxygen sensors are positioned in holes in the bottom plate to address sensor spots in the HepaChip-MP. E) The HepaChip-MP is positioned by milled guiding lines to precisely align the oxygen sensor spots with the optical fibers. F) The oxygen readout plate is compatible to use in a common incubator. Tanks mounted on the fluidic ports (Luer connectors) of the HepaChip-MP enable continuous perfusion of the cell culture chambers by gravitational driven flow (green: inlet tank; red: outlet tank) (A and B adapted from Busche et al., 2020).

Figure 2. OC of PHH cultured in the HepaChip-MP. A) In cell culture chambers containing microtissues, the OC drops over time. After re-adjusting the medium level difference between inlet and outlet after 12 hours, flow rate increases and the OC immediately rises to a higher level. B) Microscopic images of the cell culture chambers analyzed in A) at the start and the end of the measurement exhibit a slight decrease of cell number. C) Representative images of cells cultured in the HepaChip-MP from the day of seeding (day 0) and day 2 and day 3 of culture. Oxygen measurements after substance treatment were carried out at day 2 and day 3 of the culture in this study.

Figure 3. OC and ATP-levels after fructose treatment of PHH. A) While OC rapidly decreases to < 50 % of the basal OC in response to fructose treatment and reaches initial levels again after approximately 110 minutes (black graphs), glucose treatment did not substantially affect OC (grey graphs). B) Cells were treated with 20 mM lactate (left) and 20 mM glycerol (right). Neither lactate nor glycerol affected OC in comparison to control cells treated with medium only. C) The ATP level determined 20 min after the start of the fructose treatment decreased to approximately 30 % of the level of non-treated control cells. Even after reaching close-to-initial levels of cellular OC (illustrated in panel below), the ATP levels did not increase but remained at 30 % of the control cells two hours after the start of the fructose treatment (n = 3 for medium, n = 4 for 20 min and 120 min fructose; significance: **** equals p-values < 0.0001).

Figure 4. Repeated treatment of hepatocyte cultures by fructose on two consecutive days demonstrates the reproducibility of the effect of fructose on OC. The same cell culture chambers containing PHH were repeatedly treated for 2 hours with 40 mM fructose on DIV (days in vitro) 2 (A) and DIV 3 (B). Fructose treatment reproducibly induced the initial decrease in OC followed by an increase to almost initial levels (black graphs). In addition, the intensity of the relative fructose effect on OC is comparable intra-day as well as inter-day. In order to allow the cells to equilibrate to normal metabolic conditions, cells were cultured in standard culture medium for 2 hours in between fructose treatments. Cells cultured in standard medium, did not show substantial changes in OC (grey graphs).

Figure 5. OC of PHH after subsequent application of oligomycin A, FCCP and rotenone+antimycin A (R/A). A) Oligomycin A causes a drop to approximately 25 % of the initial OC. FCCP leads to a rapid increase to > 200 % of the OC, while R/A treatment results in a rapid decrease to negligible OC values. B) Cells were pretreated for 2 hours with 100 mM ammonium chloride (black graphs) or culture medium (grey graphs). Oligomycin A and R/A caused a similar decrease of the OC for pretreated as well as non-pretreated cells as observed in A). However, while the FCCP treatment caused the expected rapid increase until reaching a plateau for non-pretreated cells, cells pretreated with 100 mM ammonium chloride showed a sigmoidal-like increase of the OC thereafter (pointed out by the black ellipse).

Figure 6. Analysis of the influence of the treatment with different concentrations of diclofenac for 24 hours. A) OC after subsequent application of oligomycin A, FCCP and R/A of PHH pretreated with diclofenac. B) Detailed analysis of cellular respiration of PHH treated with different concentrations of diclofenac (n=2 for 450 µM diclofenac; n=3 for 150 µM and 50 µM diclofenac; significance: * equals p-values < 0.05).