Multiplex Analysis of 3D Liver Cell Cultures in GrowDex®

Download pdf

Johanna Lampe1, Jonathan Sheard1, Darren Heywood2, Katja Ahokas3, Piia Mikkonen1
1UPM Biomedicals, Helsinki, Finland
2Promega, Southampton, UK
3Promega, Stockholm, Sweden


Drug discovery is a lengthy and expensive process, with most drug candidates failing in the development process and never making it to market. To improve the success rate of drug development, scientists and physicians are devising novel and often complex biological tools that mimic in vivo conditions in a more reliable manner compared to the current 2D cell cultures or animal models. The increased complexity of these in vivo-like biological structures, exemplified in 3D cell cultures and engineered tissues, present a challenge for the standard cryo-transport procedures. In the standard approach, cells are exposed to at least one freeze-thaw cycle which is known to cause cell death, reduction in cell proliferative capacity and altered gene expression, as reviewed by Baust, J.G. et al., (2009) [1]. Furthermore, cells are affected by the metabolic activity and toxicity of the cryoprotectants that are integral to the cryopreservation process. 

Brain organoids are highly incompatible with cryopreservation due to the large amount of cell death that occurs for many tissue types upon undergoing a freeze thaw cycle. As organoids mature, they build complex networks of different cell types, and these networks may get compromised following cryopreservation. Cellular repair and re-growth can range from slow to impossible (in the case of postmitotic neurons, for example), and new cells may not functionally integrate into the surrounding tissue. Also, many dead or dying cells negatively affect cell signaling in their immediate surrounding, further compromising cells that may have survived freezing, ultimately driving thawed tissues away from normal homeostatic behavior. 

Therefore, this partial or complete incompatibility with cryopreservation, limits their application in high-throughput screening (HTS) or high-content screening (HCS) workflows.


  • Differentiated HepaRG (Cat No. HPR116080, Biopredic)
  • William’s E medium (Cat No. 12551032, Gibco)
  • GlutaMax (Cat No. 35050038, Gibco)
  • HepaRG Thawing/Plating/General Purpose Medium Supplement with antibiotics (Cat No. ADD670C, Biopredic)
  • HepaRG Maintenance/Metabolism Medium Supplement with antibiotics (Cat No. ADD620C,
  • PhenoPlate 96-well, black, optically clear flat-bottom, ULA-coated (Cat No. 6055802, Revvity)
  • 96 Well White/Clear Bottom Plate (Cat No. 10067581, Fisher Scientific).
  • Low-retention pipet tips 1000 µl (Cat No. 11535454, Fisher Scientific)
  • Low-retention pipet tips 200 µl (Cat No. 70.3031.275, Sarstedt)
  • GrowDex®, 1.5% (Cat No. 100 103 005, UPM Biomedicals)
  • CellTox™ Green Cytotoxicity Assay (Cat No. G8742, Promega)
  • P450-Glo CYP3A4 Assay with Luciferin-IPA (Cat No. V9002, Promega)
  • P450-Glo CYP2C9 Assay with Luciferin-H (Cat No. V8792, Promega)
  • CellTiter-Glo® 3D Cell Viability Assay (Cat No. G9682, Promega)
  • Human Albumin ELISA Kit (Cat No. ab179887, abcam)
  • Rifampin (Cat No. 15473539, Fisher Scientific)
  • Itraconazole (Cat No. S2476, Selleckchem)
  • ATP solution 100 mM (Cat No. 10304340, Thermo Fisher)
  • Victor Nivo multimode plate reader (Cat No. HH35000500, Revvity)
  • CertusFlex (Fritz Gyger AG)
  • GraphPad Prism 10.0.3 (GraphPad Software) 



  1. A 1% solution of GrowDex was prepared using the HepaRG Thawing/Plating/General Purpose Medium. Briefly, to each ml media in a reaction tube 2 ml GrowDex were added, then mixed thoroughly by stirring and pipetting whilst avoiding bubble formation.
  2. Differentiated HepaRG cells were thawed according to the manufacturer’s instructions, counted, and mixed with the 1% GrowDex solution and with media to reach 750,000 cells/ml in a 0.5% GrowDex solution. This solution was transferred to black ultra low adhesion 96-well plates at 100 µl/well, and 100 µl HepaRG Thawing/Plating/General Purpose medium were gently added on top. One plate per analysis timepoint was prepared.
  3. To prepare samples for blank/background measurements, six wells on each plate were filled with 100 µl cell-free 0.5% GrowDex in HepaRG Thawing/Plating/General Purpose Medium, and 100 µl medium were added on top.
  4. Following a 2 d incubation, medium was changed to HepaRG Maintenance/Metabolism medium containing 1:500 CellTox-Green dye. Due to the GrowDex in the wells, the final concentration was 1:1000. This and all following media changes were also performed for the wells containing GrowDex without cells.
  5. HepaRG Maintenance/Metabolism medium was renewed three times a week, then containing 1:1000 CellTox-Green dye.
  6. Before each change of medium, cell death was quantified by recording the fluorescence of the CellTox-Green dye in a plate reader with settings according to the kit instructions. The cell-free GrowDex-only wells were used for background measurement.
  7. After 7/14/21/28 d spheroid culture, the cells in one plate were treated with Rifampin or Itraconazole, or control-treated with media, all of these containing 1:1000 DMSO. The drug concentration in the supernatant was doubled compared to the final concentration, taking the 1:2 dilution by the 100 µl GrowDex into account.
  8. After 2 d drug treatment, cultures were analyzed for cell death as described above. Then, the supernatants were collected for further analysis (e.g. quantification of albumin secretion) and stored at -20 °C, and CYP activity as well as cell viability were analyzed as described below.
  9. CYP3A4 activity: The P450-Glo kit components were thawed and the Luciferin Detection Reagent was reconstituted according to the kit instructions. Then, HepaRG Maintenance/ Metabolism medium containing the CYP3A4 substrate was prepared. To reach the final dilution 1:1000 (3 µM) on the cells, the concentration in the media was tripled, meaning a dilution of 1:333.33. 50 µl of this solution were gently added on top of the cultures in GrowDex, as well as into the wells containing GrowDex only. The plates were incubated at 37 °C/5% CO2 for 2 h. Following this incubation, 25 µl of the supernatant were transferred to a new, white walled 96-well reaction plate, and 25 µl of the reconstituted Luciferin Detection Reagent were added with Certus Flex. The plate was incubated at room temperature for 20 min, and luminescence was recorded in a plate reader. The results were blank-corrected and normalized to the ATP content of the wells determined by CellTiter-Glo 3D.
  10. Cell viability: The CellTiter-Glo 3D reagent was thawed at 4 °C overnight, equilibrated at room temperature, and mixed by gently inverting before use. A dilution series of ATP standard in HepaRG Maintenance/Metabolism medium was prepared at concentrations 0.5 nM, 1 nM, 5 nM, 10 nM, 50 nM, 100 nM, 500 nM, 1 µM, 5 µM, 10 µM, 50 µM, 100 µM. 100 µl of these were added in duplicates to empty wells on the 96-well plate. The plate was equilibrated to room temperature, the leftover supernatant was removed from the cells/GrowDex, and 100 µl CellTiter-Glo 3D reagent were added to each well. After mixing for 5 min at room temperature the incubation was continued for 25 min, and the luminescence recorded according to the kit instructions. NB: extended shaking can improve the signal intensity. The ATP content of the samples was extrapolated from the ATP standard curve.
  11. Quantification of albumin secretion: The albumin content of cell culture supernatants was determined via ELISA. The supernatants were thawed, centrifuged at 2000xg for 10 min and diluted 1:200. The assay was performed according to the kit instructions. The albumin content was extrapolated from the standards, and the results were normalized to the ATP content of the wells.



HepaRG cells in GrowDex started clustering within 1-2 days after plating. After 5-7 d of culture, spheroids of 50-70 µm had formed (Fig. 1), and this size remained stable for the duration of the experiment. After 1/2/3/4 weeks, the cells were treated with a standard CYP inhibitor (Itraconazole) and inducer (Rifampin), and readouts were performed after 2 d incubation with the compounds. 

Figure 1. HepaRG spheroids in 0.5% GrowDex 9 days after plating. Scale bar 300 µm. 

Setup of multiplexing assay. Adding the CellTox-Green reagent to culture media allowed for an easy quantification of cell death at any time during the experiment without further pipetting. Therefore, this was the first assay to be performed in the multiplexing protocol, followed by P450-Glo and finally by CellTiter-Glo 3D. Supernatants of the cultures were collected after CellTox-Green measurements and before quantification of CYP activity and can be used for a multitude of different readouts, as exemplified here with quantification of albumin secretion. The P450-Glo assay was performed according to the nonlytic protocol, however, the incubation time with proluciferin had to be extended from 1 h to 2 h to obtain stronger signals that allowed more sensitive differentiation between treatments. Performing the nonlytic protocol of P450-Glo allowed for determination of cell viability via CellTiter-Glo 3D, the results of which were then used for normalization of CYP activity and albumin secretion. 

Cytotoxicity of compounds. While Itraconazole treatment after 3 weeks culture time seemed to have some cytotoxic effect, the treatments generally did not induce cell death (Fig. 2A), therefore allowing detailed analysis of hepatocyte-specific activity. This was verified by the cell viability assay CellTiter-Glo 3D. As expected for a non-dividing cell line, cell viability decreased over time, however, it was not affected by the treatments (Fig. 2B)

Figure 2. Effect of compound treatment and of long-term culture on HepaRG spheroids. Cells were cultured in GrowDex for the indicated times and compound-treated for 2 d before cytotoxicity and viability were quantified. Cyotoxicity relative to control (A), and cell viability as ATP concentration (B). N=6.


CYP3A4 activity. The basal CYP3A4 activity decreased over time as shown in figure 3A, however, it could be induced/inhibited throughout the 30 d experiment (Fig. 3B). Treatment with 10 µM Rifampin induced the activity by factors 2.48, 1.82, 3.78, 3.17 after 1, 2, 3, 4 weeks culture time, respectively. Itraconazole treatment led to activities of 0.08, 0.14, 0.23, 1.06 relative to control at the same timepoints. Even though CYP3A4 activity was not reduced by Itraconazole treatment after 4 weeks in culture, this shows that modulation of CYP enzyme activity is feasible in long-term HepaRG cultures in GrowDex. 

Figure 3. CYP3A4 activity. HepaRG cells were cultured in GrowDex for the indicated times and compound-treated for 2 d before CYP3A4 activity was quantified. Relative CYP3A4 activity of control-treated cells (A). Induction and inhibition of CYP3A4 activity shown as relative to control (B). N=6


Figure 4. Albumin secretion. HepaRG cells were cultured in GrowDex for the indicated times and compound-treated for 2 d  before supernatants were collected, stored at -20 °C, and analyzed for albumin content. Results were normalized to the ATP content of the cultures. N=6.


Albumin secretion. Albumin secretion slightly increased at day 16 compared to day 9, followed by a decrease at the later time points (Fig. 4). Treatment with CYP inducer and inhibitor did not affect the secretion of albumin.


This study shows that a 3D HepaRG hepatocyte model can be successfully maintained in GrowDex for up to 30 days. Consistent with expectations for a non-proliferative cell line, a decline in cell viability was observed over time. However, the liver cells continued to demonstrate specific functional capabilities, evidenced by the sustained activity of CYP3A4 and the secretion of albumin, although both parameters showed a reduction over time. Importantly, CYP3A4 activity can be modified by drug treatment throughout the 30 d incubation. 

This model can be analyzed by multiplexing of assays to quantify cell death, CYP activity, as well as cell viability (employing CellTox-Green Cytotoxicity Assay, P450-Glo CYP3A4 Assay, and CellTiterGlo 3D Cell Viability Assay, respectively), and additionally by supernatant-based readouts like ELISA to quantify albumin secretion. We demonstrated that these assays are fully compatible with GrowDex hydrogels, only a minor modification (extension of incubation time) of the standard protocol for P450-Glo was necessary. The capability to perform these assays simultaneously in a single experiment offers several benefits, such as generating more biologically relevant data, conserving precious samples, reducing experimental time, and minimising reagent use. It also streamlines workflows and cuts down on the time and costs associated with toxicity testing approaches. One step, the addition of the CYP detection reagent, was performed here with an automatic dispenser. The automation-friendly characteristics of GrowDex and of the assays can allow even more automation of the process, which would increase speed and reproducibility whilst reducing costs. We have also optimized the P450-Glo protocol for CYP2C9 in this model, showing the versatility of the multiplexing protocol. The liver cell model and multiplexing of assays described here can be used in a multitude of applications in the future. This includes liver toxicity studies, as well as quality control assays to evaluate batches of primary human hepatocytes.

In addition, more supernatant-based readouts can be easily added to this model without further optimization of the culture or multiplexing conditions. This would enable measurement of even more parameters depending on the scientific question.


  1. Allison, R., Guraka, A., Shawa, I.T., Gyan Tripathi, G., Moritz, W., Kermanizadeh, A. (2023). “Drug induced liver injury – a 2023 update.” J Toxicol Environ Health B Crit Rev 26(8)
  2. Yang, S., Ooka, M., Margolis, R.J., Xia, M. (2023). “Liver Three-Dimensional Cellular Models for High-Throughput Chemical Testing.” Cell Reports Methods 3(3)
Combining GrowDex® and spheroONE® for automated and highly reproducible 3D cell culture workflows
Story | 11 min

Combining GrowDex® and spheroONE® for automated and highly reproducible 3D cell culture workflows

Read more
Fantastic Event Through Collaboration!
Story | 2 min

Fantastic Event Through Collaboration!

Read more
Automation Compatible 3D High Throughput Liver Toxicity Testing in Animal Free Nanofibrillar Cellulose Hydrogels
Story | 1 min

Automation Compatible 3D High Throughput Liver Toxicity Testing in Animal Free Nanofibrillar Cellulose Hydrogels

Read more