Page 174 - IJB-8-2

Nephrotoxicity Testing with Bioprinted Renal Spheroids
after 3 days of treatment with different concentrations of cell culture medium containing a dilution of cisplatin
cisplatin. The composition of this dataset is described in the (c Cisplatin = 100 µM). The treatment was conducted at
Table S1. The network architecture and hyperparameters 37°C for 24 h. Then, the cisplatin dilution was replaced
were chosen by automatic hyperparameter optimization. by fresh medium. At this time point, we started the
The selected architecture was a VGG11 optimized with readout. During the subsequent incubation period, the
ADAM for 90 epochs. The full list of hyperparameters supernatant was sampled at different time points (0, 12,
is provided in the Supplementary File. The concept of 24, 36, 48, and 72 h post treatment), and the release of
automated treatment effect readout was adapted from LDH was quantified. The measured LDH release curves
previous studies [11,12] . for the 2 sample types are shown in Figure 1D and E.
For 2D monolayer cultures, no significant LDH release
3. Results and discussion was detected within the first 24 h (Figure 1D). After

3.1. Required 3D cell count for toxicity 36 h, the LDH release accounted for 60 ± 20% and
quantification increased to 96 ± 20% after 48 h, relative to the Lysed
Ctrl. At 72 h post-treatment, the released LDH decreased
First, the required total cell count per 3D sample was to 42 ± 23%. The decrease in measured LDH could be
defined, which enabled the quantification of toxicity with related to the degradation of the enzyme in the culture
a chemical LDH assay readout. Therefore, cell cluster supernatant. For the untreated control group (Neg. Ctrl),
arrays were fabricated, ranging from 1 to 400 spheroids the release of LDH continuously increased over time,
(Figure 1A). The cell clusters self-assembled to spheroids reaching a maximum of 10 ± 7% at 72 h. This could be
of uniform size within 4 days of incubation . Then, the explained by the fact that the 2D monolayer cultures
[8]
spheroids were lysed (Lysed Ctrl), and the amount of were fully confluent. At this point, the cells in culture
released LDH was compared to the Neg. Ctrl. (Figure 1B). need to adapt to the limited growth area and nutrient
The Lysed Ctrl samples showed an increasing amount deficiency, which could cause unspecific cell lysis and
of LDH in the supernatant in relation to an increasing LDH release. The 3D spheroid arrays showed a distinct
number of spheroids per array, as expected. In contrast, behavior compared to the 2D monolayers (Figure 1E).
the Neg. Ctrl samples showed a constant absorbance, The first increase in LDH release occurred 24 h post-
independent of the cell count. The positive to negative treatment, reaching 60 ± 32% of the Lysed Ctrl. After
signal ratio (SP/N = Lysed Ctrl/Neg. Ctrl) was calculated 48 h, the maximum was reached at 96 ± 7%. Afterward,
to determine the minimum spheroid count required the value declined to 79 ± 27% (72 h). Again, this decline
to distinguish the Lysed Ctrl. from the Neg. Ctrl. The could be explained by an impairment of the LDH activity
desired value of SP/N ≥3 was achieved for arrays with due to the progressive incubation period. The untreated
at least 100 spheroids. For these, an OD 450_100 = 0.8 ± 0.6 control shared a similar behavior with the 2D monolayers,
was measured, showing a high coefficient of variation with a slight and static linear increase of released LDH
(75%). A number of 225 spheroids showed an absorbance over time, reaching a maximum of 23 ± 1% at 72 h.
of OD 450_225 = 1.3 ± 0.4. In addition, the lowest coefficient This indicates again that within 3D spheroid cultures,
of variation was found for this format (30%). In addition, unspecific cell lysis occurred, which was not related to a
the value was in the range of the LDH Ctrl, which was toxic treatment. This finding is in good accordance with
measured to OD 450_LDH Ctrl = 1.6 ± 0.3, corresponding the previously published data . Taken together, from
[15]
to 83 ± 25%. The largest array format (400 spheroids) the experiments, the optimal time-point for the LDH-
showed an absorbance of 114 ± 47% relative to the LDH assay based readout was found to be at t treatment = 48 h.

Ctrl, however, with an increased coefficient of variation. The dynamic observation of LDH release after cisplatin
Based on these findings, spheroid arrays consisting of treatment indicated significant differences between
225 spheroids, or 34×10 cells, were selected for the 2D and 3D cell models. The faster onset of toxicity in
6
quantification of toxicity using an LDH assay. 3D models could be related to a more efficient uptake
of the toxin. In the case of cisplatin, the uptake occurs
3.2. 3D response time to toxic treatment through transport proteins located on the basal side of the

Next, a suitable time-point for the toxicity readout was cells. This side is facing toward the cisplatin containing
determined. A schematic timeline of the experimental culture medium in the case of hydrogel embedded 3D
procedure is shown in Figure 1C. All samples were spheroids, while in 2D monolayers it is in contact to the
prepared by bioprinting 4 days before a treatment with solid culture vessel surface.
cisplatin, a common nephrotoxic substance . Two types 3.3. Dose-response curves
[14]
of samples were prepared, either classical monolayers
(2D) or bioprinted arrays of 225 spheroids (3D Samples). Next, dose-dependent treatment effects were quantified
On day 4 of incubation, the medium was replaced by a to derive dose-response curves for different cell culture

166 International Journal of Bioprinting (2022)–Volume 8, Issue 2