Regulation of Ribosome Biogenesis and Protein Synthesis Controls Germline Stem Cell Differentiation

Abstract

Complex regulatory networks regulate stem cell behavior and contributions to tissue growth, repair, and homeostasis. A full understanding of the networks controlling stem cell self-renewal and differentiation, however, has not yet been realized. To systematically dissect these networks and identify their components, we performed an unbiased, transcriptome-wide in vivo RNAi screen in female Drosophila germline stem cells (GSCs). Based on characterized cellular defects, we classified 646 identified genes into phenotypic and functional groups and unveiled a comprehensive set of networks regulating GSC maintenance, survival, and differentiation. This analysis revealed an unexpected role for ribosomal assembly factors in controlling stem cell cytokinesis. Moreover, our data show that the transition from self-renewal to differentiation relies on enhanced ribosome biogenesis accompanied by increased protein synthesis. Collectively, these results detail the extensive genetic networks that control stem cell homeostasis and highlight the intricate regulation of protein synthesis during differentiation.

Figure 1. Transcriptome-wide in vivo RNAi screen: workflow and summary of results

(A) Schematic representation of the Drosophila germarium. (B) Screening crosses and workflow. For primary screening, 8171 genes were knocked down using transgenic RNAi lines and the germline-specific driver (nos-Gal4). For the secondary screen, crosses were repeated for 864 candidates. (C) Pie chart summarizing primary screen results. (D-E) Confocal images showing phenotypic categories. Ovaries were stained for Vasa (green) and 1B1 (red). (F-H) Heatmaps with phenotypic frequencies (black-yellow-red scale; 0-100%) observed for knockdowns of candidates. Phenotypic group classification, color-coded as in (I), is indicated to the right of each heatmap. (F) Phenotypic results for candidates identified solely by egg laying/hatching, (G) by both RT-qPCR and egg laying/hatching or (H) exclusively by RT-qPCR assays. (I) Pie chart summarizing secondary screen results. Scale bars, 20 μM. Related to Figures S1 and S2, and Tables S1 and S3.

Figure 2. Germline maintenance factors are enriched for highly expressed genes functioning in core cellular networks

(A) Box plots displaying the distribution of screened genes according to their ascending expression level. Genes were organized into ten equal bins, each containing 804 genes. (B) The distribution of hits found in each of the expression bins defined in (A). (C) The percentage of hits showing either ‘GSC loss’ or ‘differentiation defects’ for each expression bin defined in (A). (D) Gene networks identified through COMPLEAT analysis (Table S2). Circles represent genes and colors indicate phenotypic classes as displayed in the bottom panel. Gray lines designate protein interactions. (E) Venn diagram showing the overlap of ‘GSC loss’ genes and genes showing strong ovarian defects when knocked down in somatic ovarian cells (Handler et al., 2013). Related to Figure S3 and Table S2.

Figure 3. Knockdown of ribosome assembly genes causes ‘stem cysts’

(A) GO term analysis for genes showing ‘no germline’ or ‘differentiation’ phenotypes (Table S2). P-values are shown in white to blue scale. (B) Frequency of ‘no germline’ and ‘differentiation’ hits (shown as a percent of hits from each category; green to red scale) for the different steps leading to ribosome biogenesis. Genes were annotated using the KEGG database. (C) Confocal projections of wild-type and RNAi knockdown germaria for genes involved in ribosome assembly and translation control. For convenience, we adopted yeast or mammalian nomenclature whenever homologs were clear (Woolford and Baserga, 2013). Ovaries expressing BamP-GFP were stained for Vasa (green), 1B1 (red), and GFP (blue). Grayscale panels show 1B1. (D) Confocal projections of germaria stained for Vasa (green), phosphotyrosine (red; ring canal), and GFP (blue; BamP-GFP). Stars mark terminal filaments. Scale bars, 20 μM. Related to Figures S4 and S5, and Table S2.

Figure 4. Ribosome assembly factors and eIF4E interact positively with the ESCRT-III protein Shrb to promote GSC abscission

(A) Confocal projections of germaria stained for 1B1 (red) and phosphotyrosine (green). Stars mark terminal filaments. (B-C) Confocal projections of (B) 16-cell and (C) 32-cell egg chambers. Ovaries were stained for Phalloidin (F-actin; grey) and DAPI (blue). Grayscale panels show F-actin channel in oocytes. Red arrows indicate ring canals connecting to the oocyte – four in 16-cell egg chambers and five in 32-cell egg chambers. (D) Frequency of egg chambers containing 32 cells in genetic interaction experiments. Sibling genotypes are described. *** represents Chi-square tests with p<0.001. Scale bars, 20 μM. Related to Figure S5.

Figure 5. Knockdown of dTORC1 pathway and the cochaperone adaptors dTel2 and dTti1 affects GSC nucleolar hypertrophy and germline differentiation

(A-B) Confocal projections of germaria stained for (A) Vasa (green), 1B1 (red), and GFP (blue; BamP-GFP); or (B) Vasa (green), Fibrillarin (red, nucleolus), and GFP (blue; BamPGFP). (B) Grayscale panels show Fibrillarin. Germ cells are in green. (C) Quantification of nucleolar/nuclear volume ratio in wild-type germ cells (see experimental procedure). Results were plotted relative to GSCs. *** represents one-way analysis of variance with p<0.001 and * p<0.01. (D) Quantification of GSC nuclear and nucleolar volume in RNAi knockdowns. Results were plotted in relation to wild-type GSCs from measurement presented in figure 5C. Number of analyzed cells is presented at the top of each column. Data are represented as mean ± standard deviation. Scale bars, 20 μM.

Figure 6. Phenotypic characterization of dTel2, dTti1, and their interactors

(A) Germaria carrying the Dad-LacZ reporter were stained for Vasa (green) and β-Galactosidase (magenta). (B) Germaria were stained for pMad (yellow, GSCs), 1B1 (red), and GFP (blue; BamP-GFP). (C) Confocal projections of germaria stained for Vasa (green) and Fibrillarin (red). (D) Quantification of GSC nuclear and nucleolar volume. Results were plotted in relation to wild-type GSCs from measurements in Figure 5C. Number of analyzed cells is at the top of each column. Data are represented as mean ± standard deviation. (E) Confocal projections of whole ovaries stained for Vasa (green) and DAPI (blue). (F-G) Confocal projections of germaria stained with (F) Vasa (green), 1B1 (red), and GFP (blue; BamP-GFP); or (G) Vasa (green), Fibrillarin (red), and GFP (blue; BamP-GFP). (G) Grayscale panels show Fibrillarin. Germ cells are outlined in green. (H) Quantification of GSC nuclear and nucleolar volume in knockdowns. Results were plotted as in (D). Scale bars, (A-C,F-G) 20 μM, (E) 100 μM.

Figure 7. Protein synthesis during GSC differentiation

(A) Representative confocal image of protein synthesis analysis in wild-type germaria expressing BamP-GFP. Live ovaries were incubated with OP-Puro (red), and stained for Vasa (green) and GFP (blue). (B) Quantification of OP-Puro fluorescence intensities. Graph summarizes data for all time points (15, 30, 45, and 60 min; Figure S6A for individual analyses). *** represents one-way analysis of variance with p<0.001. (C) Representative confocal image of protein synthesis analysis in bam^Δ86 mutant ovaries. After labeling OP-Puro (red), ovaries were stained for Vasa (green). (D) Knockdown of ribosome assembly and translation initiation factors results in accumulation of germline ‘stem cysts’. (E) Knockdown of genes positively regulating the function of the transcription co-factor dTRRAP and the TOR kinase leads to the accumulation of undifferentiated cells (GSCs and CBs). Scale bars, 20 μM. Related to Figure S6.