Fig 1.
CSE-treated bronchospheres have decreased spheroid area and increased MUC5AC secretion in the lumen compared to control.
A) Schematic of a well within a 384-well plate with bronchosphere organoids in 2.5D culture with a 3T3 feeder cell layer at the well bottom. B) Schematic of the experimental workflow. Bronchospheres were seeded and differentiated over two weeks, after which they were exposed to 0 or 3% CSE for 1-, 2-, and 3-week continuous treatment. At each timepoint, spheroids were imaged live to assess changes in size, then fixed and stained for secreted mucins. C) Representative images of live bronchospheres stained with TMRM dye that show changes in area with CSE treatment. D) Bar plot of spheroid area at all timepoints treated with 0 or 3% CSE. Conditions are comprised of 3 biological replicates (3 separate experiments) from 3 independent donors (9 experiments total). Datapoints were normalized to the average of the control condition for each donor and at each timepoint, due to donor-to-donor variability. Data is represented as mean ± SEM and each timepoint was analyzed by paired two-tailed t-test, **p<0.01, ***p<0.001. E) Representative images of bronchospheres stained for MUC5AC and MUC5B that show increased MUC5AC secretion in the bronchospheres lumen with CSE treatment. F) Bar plot of ratio of MUC5AC and G) MUC5B cluster area to total spheroid area. Conditions are comprised of 3 biological replicates (3 separate experiments) from 3 independent donors (9 experiments total). Datapoints were normalized to the average of the control condition for each donor and at each timepoint, due to donor-to-donor variability. Data is represented as mean ± SEM and each timepoint was analyzed by paired two-tailed t-test, *p<0.05.
Fig 2.
CSE-treated bronchospheres capture smoke signatures in humans.
A) Schematic of RNA sequencing and bioinformatics analysis. The number of DEGs at each timepoint and pooled timepoints were calculated with FDR controlled at 5% and fold-change >2. B) Enrichment of up- and down-regulated genes from CSE-treated bronchospheres in smoking signatures comprised of 7 different comparisons generated from 3 independent datasets (a legend of the comparisons is provided in S6 Table in S1 File). The heatmap on the right indicates enrichment at each timepoint with the 7 dataset comparisons. As expected, a positive enrichment score with the UP comparisons. C) Heat map visualization of gene expression changes (β estimates) of 20 top up-regulated and 20 top down-regulated genes from Shaykhiev-3 COPD comparison against all other patient comparisons as well as bronchosphere datasets, both pooled and individual timepoints. D) GSEA shows up- and down-regulated pathways in CSE-treated bronchospheres. Significantly associated pathways with gene-set overlap size larger than 30 genes were selected for visualization.
Table 1.
Effect of CS exposure on cell type marker expression.
Fig 3.
Primary small molecule screen to identify attenuators of CSE-induced decrease in size and increase in MUC5AC.
A) Schematic of primary screen set-up, endpoints, and identification of hits. Compounds were tested at 1 and 10μM concentrations with n = 3 replicates per compound and concentration. The deck consisted of 301 compounds of diverse mechanisms of action. Hit compounds at 1 and 10μM concentrations that B) increased spheroid size (swell) or C) decreased MUC5AC staining in the lumen relative to DMSO and CSE 3%-treated bronchospheres. A detailed explanation of hit selection for both readouts is provided in the methods.
Fig 4.
Validation of primary screen hits reveals compounds that recapitulate increased spheroid size (swell) or decreased MUC5AC in the lumen.
Compounds that were identified in the primary screen as modulators of spheroid size/swell were tested at A) 1μM and B) 10μM concentrations to see if they would recapitulate the phenotype. Compound 206 (1μM) significantly increased spheroid size relative to DMSO + CSE 3% control bronchospheres. Compounds that were identified in the primary screen as modulators of MUC5AC reduction were tested at C) 1μM and D) 10μM concentrations to see if they would recapitulate the phenotype. Compounds 186 (1μM), 16 (10μM), 41(10μM), 158(10μM), 243(10μM), and 253(10μM) significantly reduced MUC5AC ration within the spheroid lumen compared to DMSO + CSE 3% control bronchospheres. Compounds that appeared to cause toxicity were identified by observing a significant decrease in spheroid counts in the swell readout or a significant decrease in the total nuclear intensity of the MUC5AC ratio readout. These data are reported in S3 Fig. Compounds that correctly attenuated the phenotype but produce the opposite phenotype in the other readout (i.e., increased spheroid area but also significantly increased MUC5AC ratio in the lumen) were also filtered out. These data are reported in S4 Fig. All individual data points represent biological replicates. All plots were analyzed by ordinary one-way ANOVA with Dunnett’s multiple comparisons test. *p<0.05; **p<0.01, **p<0.01, ***p<0.001, ****p<0.0001.
Fig 5.
Attenuation of spheroid size decrease or MUC5AC increase by hit compounds.
Final hit compounds (with hit concentration and corresponding target) that A) increased bronchosphere size compared to DMSO + 3% CSE control bronchospheres or B) decreased secreted MUC5AC. Mucus reduction hit compounds were further refined by manually comparing TMRM live dye-stained images of compound treated vs. control bronchospheres to observe aberrant phenotype that could indicate an unhealthy state. These images are shown in S5 Fig.
Table 2.
Final list of hit compounds from the validation screen and their corresponding targets.