Automated high-content imaging in iPSC-derived neuronal progenitors

Abstract

Induced pluripotent stem cells (iPSCs) have great potential as physiological disease models for human disorders where access to primary cells is difficult, such as neurons. In recent years, many protocols have been developed for the generation of iPSCs and the differentiation into specialised cell subtypes of interest. More recently, these models have been modified to allow large-scale phenotyping and high-content screening of small molecule compounds in iPSC-derived neuronal cells. Here, we describe the automated seeding of day 11 ventral midbrain progenitor cells into 96-well plates, administration of compounds, automated staining for immunofluorescence, the acquisition of images on a high-content screening platform and workflows for image analysis.

Keywords: High-content screening; Image analysis; Immunofluorescence; Lab automation; Multi-well plates; Neurons.

Fig. 1

Overview of mDA progenitor screening protocol. A. mDA neuronal differentiation. Basal media and patterning factors required for each stage of the process are shown. On Day 11, compounds are dispensed from the source plates onto PO/FN/LN-coated, dried, plates. Subsequently, MDa progenitors are plated and cultured for another 24 h prior to fixing and staining. B. Image Analysis. Representative images. LC3 puncta and nuclei are counted for each imaged field, and LC3 puncta / nuclei values are calculated. Through image segmentation, LC3 puncta outside the cytoplasm (background noise) are excluded from the analysis. Abbreviations: EBs: embryoid bodies, FN: fibronectin, LN: laminin, mDA: midbrain dopaminergic, PO: poly-L-ornithin.

Fig. 2

Characterization of ventral midbrain progenitors at Day 11 of differentiation. A. Immunofluorescence. Expression of ventral midbrain markers forkhead box protein A2 (FOXA2) and homeobox transcription factor 1 alpha (LMX1A) at Day 11 of differentiation. Co-expression of FOXA2 and LMX1A is thought to be unique to the ventral midbrain, hence a high percentage of doubly-positive cells indicates effective initiation of differentiation into mDA neurons. Quantification via counting doubly-positive cells (e.g., on ImageJ) is recommended; the relevant quantification graph suggests effective differentiation into ventral midbrain progenitors as per relevant literature [17,18]. B. Real-Time Quantitative Reverse Transcription PCR (qRT PCR). Day 11ventral midbrain progenitors tested for appropriate gene expression suggestive of effective initiation of mDA neuronal differentiation. For each marker/ target, $\Delta C_T$mean values on Day 11 (calculated using the target's $C_T$ mean values, normalised versus the ones from the housekeeping gene GAPDH) are compared to ones in corresponding iPSCs. The line exhibits downregulation in gene expression of pluripotency-related markers OCT4 and NANOG, as well as upregulation in gene expression of ventral midbrain markers FOXA2, LMX1A, LMX1B, EN1 and EN2 . When such immunofluorescence marker ratios and qRT PCR marker patterns are achieved, one can be confident of effective generation of ventral midbrain progenitors; if the differentiation shows different patterns, cells must be discarded, and new differentiations started.

Fig. 3

Automated compound delivery. A. Schematic drawing of the 384 LDV Echo source plate containing the first 80 individual drugs of a compound library. The library can be dispensed into 96 well assay plates using the Labcyte Echo 550 Acoustic dispenser. B. Final plate layout in the 96 well assay plate containing controls and library compounds.

Fig. 4

Assay reproducibility and capacity for high content. A. To assess reproducibility, a pilot source plate was created containing i) compounds known to enhance LC3 puncta production (either by inducing autophagy flux or blocking autophagosome to lysosome fusion), ii) compounds with no expected effect on LC3 puncta, as well as iii) negative and positive controls (DMSO and Bafilomycin A1 in columns 1 and 12, respectively). After treatment with these compounds, destination plates in technical triplicates were fixed, stained and imaged. Image analysis generated heatmaps depicting increasing LC3 puncta per nuclei numbers in darker shades of orange. B. Compound treatments reproducibly generated similar numbers of LC3 puncta per nuclei in all technical replicates. In detail, we observed standard deviation for the 3 replicates ranging from 0.028 and 0.592, with mean and median values of 0.209 and 0.157 respectively (attesting to the reproducibility of the assay). C. The assay can be used to image multiple stains simultaneously [e.g., nuclear stains (DAPI), mitochondrial markers (HSP60), autophagy (LC3) and lysosomal (LAMP1) markers] and their response to different treatments, hence providing multiple parameters that can be useful for secondary hit validation. Apart from puncta numbers, multiple other parameters can also be analysed, such as cell morphology, spot area, intensity and marker co-localization.