Skip to main content
. 2020 Nov 3;9:e52904. doi: 10.7554/eLife.52904

Figure 8. Automated midbrain organoids possess functional characteristics of midbrain tissue and allow assessment of neural subpopulations for high-throughput screening.

(a/b) The combination of AMOs and our automated whole mount staining and clearing workflow allows the quantification of dopaminergic neuron-specific toxicity in 3D. 6-Hydroxy dopamine and MPP+ specifically ablate TH-positive dopaminergic neurons from the AMOs in a dose-dependent manner and with little variation between replicates and cell lines. n ≥ 6 organoids per data point, Error bars: SEM. Organoids at day 56 of differentiation. (c) 35 days old AMOs secrete dopamine into their cell culture medium under standard culture conditions and without further stimulation, as confirmed by ELISA. The concentration is in the same range as the dopamine levels measured in the cerebrospinal fluid (CSF) of healthy, adult humans as reported by Goldstein et al., 2012. nLine 1 = 4, nLine 2 = 3, nhCSF = 38, Error bars: SEM. (d) 70–80% percent of cells within the AMOs are negative for the precursor marker Sox2 after 30 days of differentiation. AMO line 1: nBatch1 = 90, nBatch2 = 16, nBatch3 = 15; AMO line 2: nBatch1 = 89, nBatch2 = 16, nBatch3 = 14; Error bars: SD. (e–j) AMOs respond most strongly to dopaminergic modulation and, to a lesser extent, also to glutamatergic modulation while the automated cortical hiPSC organoids are mostly affected by compounds targeting glutamatergic neurons. MEA measurements of individual AMOs (e–g) or cortical hiPSC organoids (h–j) were performed in three stages on the same sample: first, under basal conditions (black line), second, after treatment with an agonist (blue line), and third, after addition of an antagonist (red line). The pharmacological modulators targeted dopaminergic (e/h), GABAergic (f/i), or glutamatergic (g/j) neurons. The gaps in the X-axis represent the addition of the different compounds and the time we allowed for the solution to equilibrate. Shown is the raw signal of one representative example of n = 4. AMOs were at 33 days and hiPSCs at 35 days of differentiation. (k) Quantification of the effects of pharmacological modulation on AMOs and automated hIPSC organoids as measured by MEA. The bar graph shows the sum of the absolute electric field potential oscillations over 15 s of time relative to basal conditions for each modulator. Each bar represents the mean +/- SEM for n = 4 replicates, one representative raw measurement per condition is shown in (e-j). n = 4, except nhiPSC organoids Glutamate = 3; Also see Figure 8—figure supplement 1.

Figure 8.

Figure 8—figure supplement 1. AMOs allow HTS-compatible toxicity evaluation in whole organoids or specific cellular subpopulations.

Figure 8—figure supplement 1.

(a) Higher toxin concentrations resulted in increased cCasp3+ apoptotic signal in AMOs. Panel shows representative single plane confocal micrographs from our high-content analysis pipeline after adding G418 for 4 days starting at day 50 of culture. Scale bars = 100 μm. (b) Increase of cCasp3+ cells follows a typical sigmoidal dose-response kinetic in AMOs with escalating concentrations of G418. Depicted is the total number of cCasp3 cells per organoid normalized by the organoid area on the y-axis against the logarithmic concentration of G418 on the x-axis. n ≥ 3, error bars = SEM. (c) The cCasp3 signal shows little colocalization with Sox2, indicating that not neuronal precursors but other, more mature cell types are primarily affected by the treatment. The percentage of Sox2+ neural precursors among the apoptotic cells increased with the inhibitor concentration but remained relatively low with a maximum of approximately 15%. n ≥ 3, error bars = SEM (d/e) The dose-response curve (d) and colocalization analysis (e) based on a single plane display almost identical results to the ones based on entire organoids (b/c). This potentially allows a substantial reduction of imaging time and costs by acquiring only single medial slices instead of entire organoid volumes. n ≥ 3, error bars = SEM. (f) Bar graph showing the number of apoptotic cCasp3+ cells (normalized by area) for single confocal planes. Each bar represents one plane; bars are grouped by the concentration of the G418 treatment. n ≥ 3, error bars = SEM. (g/h) Followup high-content imaging-based evaluation of G418 toxicity with two different AMO lines and hiPSC organoids using a wider range of G418 concentrations than first tested in (a–f) illustrating the drawbacks of cCasp3-based toxicity screening. As cCasp3 labels cells only for a transient time during cell death, the peak signal can be missed and results can be misleading. Here, higher concentrations of G418 resulted in lower levels of cCasp3 positive cells after 96 hr (g). The size (area of the cross section) of the organoids from (g) did not change with increased concentrations of G418 (h). n ≥ 3, Error bars: SEM. (i) ATPglo-based cell viability measurements of G418 treated AMOs and hiPSC organoids (each from the same batch and treated together with the samples shown in (g)), confirmed that higher concentrations of G418 are indeed more toxic and the cCasp3-based readout failed to detect this effect. In all three assays (g-i) AMOs showed lower variance than hiPSC organoids. nAMO line 1&2 = 6, nhiPSC organoids ≥ 5, Error bars: SD. All samples in (g-i) were at day 35 at the time of analysis.