Wang Z et al. (2018), TRRAP is a central regulator of human multicili...

XB-ART-54695

J Cell Biol 2018 Jun 04;2176:1941-1955. doi: 10.1083/jcb.201706106.

Show Gene links Show Anatomy links

TRRAP is a central regulator of human multiciliated cell formation.

Wang Z , Plasschaert LW , Aryal S , Renaud NA , Yang Z , Choo-Wing R , Pessotti AD , Kirkpatrick ND , Cochran NR , Carbone W , Maher R , Lindeman A , Russ C , Reece-Hoyes J , McAllister G , Hoffman GR , Roma G , Jaffe AB .

Abstract
The multiciliated cell (MCC) is an evolutionarily conserved cell type, which in vertebrates functions to promote directional fluid flow across epithelial tissues. In the conducting airway, MCCs are generated by basal stem/progenitor cells and act in concert with secretory cells to perform mucociliary clearance to expel pathogens from the lung. Studies in multiple systems, including Xenopus laevis epidermis, murine trachea, and zebrafish kidney, have uncovered a transcriptional network that regulates multiple steps of multiciliogenesis, ultimately leading to an MCC with hundreds of motile cilia extended from their apical surface, which beat in a coordinated fashion. Here, we used a pool-based short hairpin RNA screening approach and identified TRRAP, an essential component of multiple histone acetyltransferase complexes, as a central regulator of MCC formation. Using a combination of immunofluorescence, signaling pathway modulation, and genomic approaches, we show that (a) TRRAP acts downstream of the Notch2-mediated basal progenitor cell fate decision and upstream of Multicilin to control MCC differentiation; and (b) TRRAP binds to the promoters and regulates the expression of a network of genes involved in MCC differentiation and function, including several genes associated with human ciliopathies.

PubMed ID: 29588376
PMC ID: PMC5987713
Article link: J Cell Biol

Species referenced: Xenopus laevis
Genes referenced: ccno foxj1.2 itga6 mcc mcidas muc5b myb notch2 pcbd1 pcnt rfx3 tp63 trrap

Article Images: [+] show captions

	Fig 1 A pooled shRNA screen to identify regulators of basal cell fate. (A) Workflow of the pooled shRNA screen. Human airway basal cells were transduced with a lentiviral shRNA library 1 h after they were seeded on Transwell filters at day 0. At day 21, cells were fixed, stained, and sorted into three different groups based on expression of the ciliated cell marker FOXJ1 and the basal cell marker ITGA6. The extracted gDNA from the sorted samples was analyzed for shRNA hairpin counts by NGS. (B) Gene-centric visualization of the log2 fold change and RSA score in the ciliated (ITGA6− FOXJ1+) versus secretory (ITGA6− FOXJ1−) cells. (C) Ratio of individual shRNA hairpin counts in ciliated/secretory cells from the pooled shRNA screen. KDM8 is shown as an example of a gene whose shRNA barcode counts do not differ between ciliated and secretory cells. (D) In a primary validation, the cells from two independent donors were transduced with four shRNAs for each gene candidate and sorted with FOXJ1 and ITGA6 signals. The ratios of ciliated/secretory cells from all shRNA-treated samples were normalized to those of control cells transduced with shNT. Data were acquired from cells of two donors; mean ± SEM; ***, P < 0.001; Student’s two-tailed t test.
	Fig 2 TRRAP is required for MCC formation. (A) Schematic workflow for validation of TRRAP. Six Transwell filters of fully differentiated ALI cultures derived from cells transduced with either shNT or TRRAP shRNA (shTRRAP) were pooled together. Two thirds were fixed and stained for ciliated (FOXJ1) and basal (ITGA6) markers and analyzed by flow cytometry. The remaining third was sorted with RFP to estimate gene knockdown by qPCR. (B) Representative FACS plots from TRRAP shRNA treatment. Human airway basal cells transduced with viruses for either shNT or TRRAP shRNAs were fixed and sorted with FOXJ1 and ITGA6 signals. (C) Ratios of ciliated/secretory cells in the TRRAP shRNA–treated samples were normalized to the ones from NT controls. (D) Quantification of TRRAP knockdown. CT values from qPCR in the TRRAP shRNA transduced cells were normalized to those from shNT controls (n = 6 from two independent donors; mean ± SEM; *, P < 0.001). (E) The cells transduced with either shNT or shTRRAP were immunostained for markers of goblet cells (MUC5AC, MUC5B), club cells (SCGB1A1), ciliated cells (AC-TUBULIN, FOXJ1), or basal cells (p63). Representative images from ALI cultures immunostained for AC-TUBULIN/MUC5B (top) or AC-TUBULIN/SCGB1A1 (bottom) are shown. Bar, 10 µm. (F) The RFP+ cells expressing the indicated cell markers were normalized to the total RFP+ cells to calculate the percentage of goblet, club, ciliated, or basal cells after TRRAP knockdown or in shNT controls. n = 6 from two independent donors; mean ± SEM; , P < 0.01; ***, P < 0.001; Student’s two-tailed t test.
	Fig 3 TRRAP is enriched in ciliated cells and accumulates before FOXJ1. (A–D) During the course of differentiation at ALI, cultures were stained with antibodies to TRRAP, FOXJ1 (ciliated cells), MUC5B (goblet cells), SCGB1A1 (club cells), and p63 (basal cells). Colocalization was quantified at days 0, 3, 7, and 14 at ALI (A′′′–D′′′). At ALI day 14, TRRAP colocalizes with FOXJ1 in ciliated cells (arrows in A–A′′) but rarely accumulates in other cell types (A′–D′′). (E–G) At day 0, before culture at ALI, TRRAP has begun to accumulate in nuclei with a small subset of those being coenriched for FOXJ1 (arrows; A′′′; arrow in E–E′′). At days 3 and 7 of culture at ALI, the number of cells enriched for both TRRAP and FOXJ1 increases (A′′′; arrows in F–G′′), and the number of TRRAP-only cells decreases. Notably, cells with high TRRAP accumulation have low FOXJ1 (F–G′′, arrowheads), whereas cells with high FOXJ1 accumulation have low TRRAP (F–G′′, arrows), suggesting that their expression is temporally distributed in the same cell type. Bars, 20 µm.
	Fig 4 TRRAP acts downstream of Notch2 signaling to regulate MCC formation. (A) Representative FACS plots of shRNA-infected ALI cultures that were treated with IgG or α-N2. Human airway basal cells were transduced with either shNT or shTRRAP-a at day 0. The cells were cultured in medium containing either 1 µg/ml IgG or α-N2 from day 7 (ALI day 0). At day 21, cells were stained for FOXJ1 and ITGA6 to label ciliated and basal cells, respectively, and analyzed by flow cytometry. (B) The ratio of ciliated/secretory cells of each sample was normalized to the control cells treated with IgG and shNT (n = 6 from two independent donors; mean ± SEM). (C) Representative images of ALI cultures with the indicated treatments, stained for AC-TUBULIN, to label ciliated cells, and either MUC5B (top row) or SCGB1A1 (bottom row) to label goblet or club cells, respectively. Bar, 10 µm. (D) Quantification of the percentage of RFP+ (shRNA-containing) ciliated (AC-TUBULIN+), goblet (MUC5B+), and club (SCGB1A1+) cells. Data were obtained from six independent cultures with two independent donors. Error bars indicate SEM. , P < 0.05; , P < 0.01; **, P < 0.001; Student’s two-tailed t test.
	Fig 5 TRRAP regulates MCC formation upstream of MULTICILIN. (A) Human airway basal cells transduced with either shNT or shTRRAP-b at day 0 were differentiated at ALI and stained for RFP (red), PERICENTRIN (green), or MUC5B (white). Arrowheads highlight RFP+MUC5B− presumptive ciliated cells. Note that knockdown of TRRAP results in an inhibition of centriole replication. Bar, 10 µm. (B) Quantification of the percentage of shRNA-expressing presumptive ciliated cells (RFP+ MUC5B−) with more than two centrioles. Each condition was performed in two independent donors, and the total number of cells quantified is indicated. (C) Representative FACS plots of day 21 ALI cultures, which had been transduced with lentivirus encoding for GFP or MULTICILIN-GFP, with either shNT or shTRRAP-b at day 0. Cultures were dissociated and then sorted based on FOXJ1 and ITGA6 signals. (D) Quantification of the ratio of ciliated/secretory cells from the indicated treatments, normalized to control cells that were cotransduced with shNT and GFP (n = 3). Data shown are represented as mean ± SEM; , P < 0.05; *, P < 0.01; Student’s two-tailed t test.
	Fig 6 TRRAP regulates a network of genes controlling multiple steps of multiciliogenesis. (A) Schematic of the experimental design. (B) Venn diagram of the number of TRRAP DEGs, identified by RNA-seq, and the number of promoter regions (<1 kb from the transcription start site) containing TRRAP peaks. (C) Enrichment analysis of TRRAP DTGs in the indicated datasets. The number of DTGs within each dataset is indicated. Shown are the p-values calculated by Fisher’s exact test with Bonferroni correction. Note the TRRAP DTG enrichment in sets of MCC genes (Hoh et al., 2012; van Dam et al., 2013; Nemajerova et al., 2016; Quigley and Kintner, 2017) and genes that are mutated in PCD; (Horani and Ferkol, 2016) but not in the set of basal cell genes (Hackett et al., 2011). (D) Hierarchical clustering of the 271 ciliogenesis DEGs at ALI days 4 and 14 in shNT, shTRRAP-a, and shTRRAP-b samples. Shown are the z-scores for each gene across all samples. Right: Bar plot displaying the maximum log2 expression differences between shTRRAP and shNT samples. Gray, DEGs; blue, DEGs containing a TRRAP peak in the promoter. (E) ChIP-seq tracks displaying TRRAP binding at the MCIDAS, CCNO, MYB, and RFX3 genomic loci. Shown is the fold enrichment over input.
	Figure 1. A pooled shRNA screen to identify regulators of basal cell fate. (A) Workflow of the pooled shRNA screen. Human airway basal cells were transduced with a lentiviral shRNA library 1 h after they were seeded on Transwell filters at day 0. At day 21, cells were fixed, stained, and sorted into three different groups based on expression of the ciliated cell marker FOXJ1 and the basal cell marker ITGA6. The extracted gDNA from the sorted samples was analyzed for shRNA hairpin counts by NGS. (B) Gene-centric visualization of the log2 fold change and RSA score in the ciliated (ITGA6− FOXJ1+) versus secretory (ITGA6− FOXJ1−) cells. (C) Ratio of individual shRNA hairpin counts in ciliated/secretory cells from the pooled shRNA screen. KDM8 is shown as an example of a gene whose shRNA barcode counts do not differ between ciliated and secretory cells. (D) In a primary validation, the cells from two independent donors were transduced with four shRNAs for each gene candidate and sorted with FOXJ1 and ITGA6 signals. The ratios of ciliated/secretory cells from all shRNA-treated samples were normalized to those of control cells transduced with shNT. Data were acquired from cells of two donors; mean ± SEM; ***, P < 0.001; Student’s two-tailed t test.
	Figure 2. TRRAP is required for MCC formation. (A) Schematic workflow for validation of TRRAP. Six Transwell filters of fully differentiated ALI cultures derived from cells transduced with either shNT or TRRAP shRNA (shTRRAP) were pooled together. Two thirds were fixed and stained for ciliated (FOXJ1) and basal (ITGA6) markers and analyzed by flow cytometry. The remaining third was sorted with RFP to estimate gene knockdown by qPCR. (B) Representative FACS plots from TRRAP shRNA treatment. Human airway basal cells transduced with viruses for either shNT or TRRAP shRNAs were fixed and sorted with FOXJ1 and ITGA6 signals. (C) Ratios of ciliated/secretory cells in the TRRAP shRNA–treated samples were normalized to the ones from NT controls. (D) Quantification of TRRAP knockdown. CT values from qPCR in the TRRAP shRNA transduced cells were normalized to those from shNT controls (n = 6 from two independent donors; mean ± SEM; *, P < 0.001). (E) The cells transduced with either shNT or shTRRAP were immunostained for markers of goblet cells (MUC5AC, MUC5B), club cells (SCGB1A1), ciliated cells (AC-TUBULIN, FOXJ1), or basal cells (p63). Representative images from ALI cultures immunostained for AC-TUBULIN/MUC5B (top) or AC-TUBULIN/SCGB1A1 (bottom) are shown. Bar, 10 µm. (F) The RFP+ cells expressing the indicated cell markers were normalized to the total RFP+ cells to calculate the percentage of goblet, club, ciliated, or basal cells after TRRAP knockdown or in shNT controls. n = 6 from two independent donors; mean ± SEM; , P < 0.01; ***, P < 0.001; Student’s two-tailed t test.
	Figure 3. TRRAP is enriched in ciliated cells and accumulates before FOXJ1. (A–D) During the course of differentiation at ALI, cultures were stained with antibodies to TRRAP, FOXJ1 (ciliated cells), MUC5B (goblet cells), SCGB1A1 (club cells), and p63 (basal cells). Colocalization was quantified at days 0, 3, 7, and 14 at ALI (A′′′–D′′′). At ALI day 14, TRRAP colocalizes with FOXJ1 in ciliated cells (arrows in A–A′′) but rarely accumulates in other cell types (A′–D′′). (E–G) At day 0, before culture at ALI, TRRAP has begun to accumulate in nuclei with a small subset of those being coenriched for FOXJ1 (arrows; A′′′; arrow in E–E′′). At days 3 and 7 of culture at ALI, the number of cells enriched for both TRRAP and FOXJ1 increases (A′′′; arrows in F–G′′), and the number of TRRAP-only cells decreases. Notably, cells with high TRRAP accumulation have low FOXJ1 (F–G′′, arrowheads), whereas cells with high FOXJ1 accumulation have low TRRAP (F–G′′, arrows), suggesting that their expression is temporally distributed in the same cell type. Bars, 20 µm.
	Figure 4. TRRAP acts downstream of Notch2 signaling to regulate MCC formation. (A) Representative FACS plots of shRNA-infected ALI cultures that were treated with IgG or α-N2. Human airway basal cells were transduced with either shNT or shTRRAP-a at day 0. The cells were cultured in medium containing either 1 µg/ml IgG or α-N2 from day 7 (ALI day 0). At day 21, cells were stained for FOXJ1 and ITGA6 to label ciliated and basal cells, respectively, and analyzed by flow cytometry. (B) The ratio of ciliated/secretory cells of each sample was normalized to the control cells treated with IgG and shNT (n = 6 from two independent donors; mean ± SEM). (C) Representative images of ALI cultures with the indicated treatments, stained for AC-TUBULIN, to label ciliated cells, and either MUC5B (top row) or SCGB1A1 (bottom row) to label goblet or club cells, respectively. Bar, 10 µm. (D) Quantification of the percentage of RFP+ (shRNA-containing) ciliated (AC-TUBULIN+), goblet (MUC5B+), and club (SCGB1A1+) cells. Data were obtained from six independent cultures with two independent donors. Error bars indicate SEM. , P < 0.05; , P < 0.01; **, P < 0.001; Student’s two-tailed t test.
	Figure 5. TRRAP regulates MCC formation upstream of MULTICILIN. (A) Human airway basal cells transduced with either shNT or shTRRAP-b at day 0 were differentiated at ALI and stained for RFP (red), PERICENTRIN (green), or MUC5B (white). Arrowheads highlight RFP+MUC5B− presumptive ciliated cells. Note that knockdown of TRRAP results in an inhibition of centriole replication. Bar, 10 µm. (B) Quantification of the percentage of shRNA-expressing presumptive ciliated cells (RFP+ MUC5B−) with more than two centrioles. Each condition was performed in two independent donors, and the total number of cells quantified is indicated. (C) Representative FACS plots of day 21 ALI cultures, which had been transduced with lentivirus encoding for GFP or MULTICILIN-GFP, with either shNT or shTRRAP-b at day 0. Cultures were dissociated and then sorted based on FOXJ1 and ITGA6 signals. (D) Quantification of the ratio of ciliated/secretory cells from the indicated treatments, normalized to control cells that were cotransduced with shNT and GFP (n = 3). Data shown are represented as mean ± SEM; , P < 0.05; *, P < 0.01; Student’s two-tailed t test.
	Figure 6. TRRAP regulates a network of genes controlling multiple steps of multiciliogenesis. (A) Schematic of the experimental design. (B) Venn diagram of the number of TRRAP DEGs, identified by RNA-seq, and the number of promoter regions (<1 kb from the transcription start site) containing TRRAP peaks. (C) Enrichment analysis of TRRAP DTGs in the indicated datasets. The number of DTGs within each dataset is indicated. Shown are the p-values calculated by Fisher’s exact test with Bonferroni correction. Note the TRRAP DTG enrichment in sets of MCC genes (Hoh et al., 2012; van Dam et al., 2013; Nemajerova et al., 2016; Quigley and Kintner, 2017) and genes that are mutated in PCD; (Horani and Ferkol, 2016) but not in the set of basal cell genes (Hackett et al., 2011). (D) Hierarchical clustering of the 271 ciliogenesis DEGs at ALI days 4 and 14 in shNT, shTRRAP-a, and shTRRAP-b samples. Shown are the z-scores for each gene across all samples. Right: Bar plot displaying the maximum log2 expression differences between shTRRAP and shNT samples. Gray, DEGs; blue, DEGs containing a TRRAP peak in the promoter. (E) ChIP-seq tracks displaying TRRAP binding at the MCIDAS, CCNO, MYB, and RFX3 genomic loci. Shown is the fold enrichment over input.

References [+] :

Boon, MCIDAS mutations result in a mucociliary clearance disorder with reduced generation of multiple motile cilia. 2016, Pubmed, Xenbase