A new platform for high-throughput therapy testing on iPSC-derived lung progenitor cells from cystic fibrosis patients

Abstract

For those people with cystic fibrosis carrying rare CFTR mutations not responding to currently available therapies, there is an unmet need for relevant tissue models for therapy development. Here, we describe a new testing platform that employs patient-specific induced pluripotent stem cells (iPSCs) differentiated to lung progenitor cells that can be studied using a dynamic, high-throughput fluorescence-based assay of CFTR channel activity. Our proof-of-concept studies support the potential use of this platform, together with a Canadian bioresource that contains iPSC lines and matched nasal cultures from people with rare mutations, to advance patient-oriented therapy development. Interventions identified in the high-throughput, stem cell-based model and validated in primary nasal cultures from the same person have the potential to be advanced as therapies.

Keywords: CF-causing nonsense mutations; apical chloride conductance assay; complementary assays of primary and iPSC derived tissues; high-throughput phenotypic platform; pluripotent stem cells; precision medicine; therapy testing.

Figure 1

Differentiation of patient-derived iPSCs to lung progenitor cells (A) Schematic of differentiation protocol and timeline. Human iPSCs were directed to definitive endoderm and passaged onto 96 well plates during the anterior foregut endoderm differentiation. The submerged cultures were differentiated for 10 more days into lung progenitor cells (stage 3b; Wong et al., 2015). (B) Immunofluorescence images of submerged cultures. Scale bar, 80 μm. Most cells stained positive for TTF1 (NKX2-1) and SOX9. Negative controls are shown in Figure S1. (C) Principal-component analysis (PCA) comparing iPSC lines, submerged lung progenitor cultures differentiated from iPSC lines, and primary bronchial cultures. Both CF and non-CF (including mutation-corrected) iPSC lines were studied. (D) Heatmap of gene expression clustered according to cell type using marker genes (Deprez et al., 2020). The columns correspond to different donors and whether lines are CF or non-CF (including mutation corrected [MC]). Columns are also clustered as an embryonic stem cell line (H1), iPSC lines, lung progenitors differentiated from iPSC lines as submerged cultures (SC), and primary human bronchial epithelial (HBE) cultures. Relative CFTR expression across cultures is shown in the bottom row of the heatmap.

Figure 2

Lung progenitor cultures express WT CFTR and channel activity in fluorescence-based assay (A) Representative FLIPR trace (mean ± SEM) responses of four wells plated with submerged cultures stimulated by 10 μM forskolin and inhibited by 10 μM CFTRInh-172. This cell line was derived from donor CF2MC for which the F508del mutation was corrected to WT. The method for mutation editing is described in Eckford et al. (2019), and characterization of the CF2MC line is given in Document S1. The naming of lines is consistent with Figure 1. (B) Western blot shows molecular weight (MW) markers and mature CFTR expression in submerged culture of the same line, (180 kDa). Calnexin (CNX) was used as loading control. (C) Top, heatmap shows representative data for a single plate (48 wells), used for establishing assay statistics (p value, Z-prime factor, and SSMD). Multiple wells were seeded with iPSCs differentiated to submerged lung progenitor cells. Alternating rows (each containing six wells) show well scans of CFTR channel activation after agonist (forskolin, FSK) or vehicle control (DMSO) addition. The response size is color coded as shown in the side bar, with red representing the highest response and purple-black, the lowest response. Bottom, bars show data from all of the above wells treated with control (DMSO) or FSK. Assay statistics are superimposed.

Figure 3

Submerged lung progenitor cultures generated from iPSCs from two different donors homozygous for F508del exhibit robust responses to modulators (A) Left: representative FLIPR traces of cultures derived from iPSCs with F508del mutation after chronic rescue (48 h) with DMSO (0.1%) or small molecules as defined in the key with concentrations indicated in Table S1. After a 5 min baseline, the cells were stimulated with DMSO or FSK ± 1 μM VX-770 (or 1.5 μM AP2). Right: bar graph shows peak response from each modulator combination. Each solid circle represents mean peak response of 4 wells in a 96 well plate for four plates of lung progenitor culture generated from a single differentiation of iPSC line CF2. The horizontal line in each bar, indicates the range amongst the mean measurements. The naming of cell lines is consistent with Figure 1. (B) Left: FLIPR traces showing responses to small-molecule modulators on epithelium differentiated from an iPSC line derived from donor CF4. Right: bars show reproducibility of the FLIPR assay. Each solid dot represents the mean peak response for 4 technical replicates (wells) of a 96 well plate, and there were four plates generated from a single differentiation of iPSC line CF4. The horizontal line in each bar, indicates the range amongst the mean measurements. Significant differences between treatment groups were determined using an ordinary one-way ANOVA, multiple comparisons test (Prism version 9.2). (C) Correlation between mean donor-specific activations measured using FLIPR (mean values ± interventions, extracted from bars above) and mean donor-specific changes measured in the Ussing chamber, delta Ieq (μA/cm²), after forskolin and treatment as reported in Laselva et al. (2018).

Figure 4

Lung progenitor cultures generated from iPSCs from three different donors with the nonsense mutation W1282X exhibit differential phenotypic responses to modulators (A) Heatmap of peak responses generated from non-CF culture stimulated with 10 μM forskolin (FSK) (left) and CF culture (W1282X) treated with a combination of small molecules (A–H, concentrations defined in Table S1; see also Video S1 for movie showing dynamic CFTR activation in the assay). (B) Bar graph showing the peak response of W1282X CFTR treated with DMSO (0.1%) or small molecules (defined on y axis with concentrations indicated in Table S1). After a 5 min baseline, the cells were stimulated with DMSO or FSK ± 1 μM VX-770 (VX) or 1.5 μM AP2 (AC). Mean peak responses (of 4 wells in a 96 well plate) after agonist and potentiator are shown for each small-molecule combination as a solid circle. The horizontal line in each bar, indicates the range amongst the mean measurements. Three patients homozygous for W1282X were studied (CF5, 6, and 7). For CF5, 5 plates generated from four differentiations from iPSCs were studied. For CF6, 3 plates from one differentiation were studied, and for CF7, 5 plates from two differentiations were studied. CFTR modulators (VX-809 or AC1 + AC2-2), in combination with SMG1i, were effective in increasing the abundance of the truncated W1282X protein in lung progenitor cultures (Figure S4). (C) Correlation plot between FLIPR peak response and Ussing chamber studies (data from Laselva et al., 2020a). (D) Basal W1282X-CFTR transcript (left) expression in primary nasal epithelial cultures (nasal) and submerged iPSC-derived lung progenitor cultures (SC) from three donors. The vertical line in each bar shows the range in CFTR expression amongst the three donors.

Figure 5

Open access bioresource (CFIT, https://lab.research.sickkids.ca/cfit/cystic-fibrosis-patients-families-researchers/cell-resources-available/) contains matched iPSCs and nasal epithelial cultures from multiple people with the same CFTR genotype to enable two-step preclinical trials that account for individual variation and select the best intervention for each individual with a rare CF-causing mutation.