Interchromosomal interaction of homologous Stat92E alleles regulates transcriptional switch during stem-cell differentiation

Abstract

Pairing of homologous chromosomes in somatic cells provides the opportunity of interchromosomal interaction between homologous gene regions. In the Drosophila male germline, the Stat92E gene is highly expressed in a germline stem cell (GSC) and gradually downregulated during the differentiation. Here we show that the pairing of Stat92E is always tight in GSCs and immediately loosened in differentiating daughter cells, gonialblasts (GBs). Disturbance of Stat92E pairing by relocation of one locus to another chromosome or by knockdown of global pairing/anti-pairing factors both result in a failure of Stat92E downregulation, suggesting that the pairing is required for the decline in transcription. Furthermore, the Stat92E enhancer, but not its transcription, is required for the change in pairing state, indicating that pairing is not a consequence of transcriptional changes. Finally, we show that the change in Stat92E pairing is dependent on asymmetric histone inheritance during the asymmetric division of GSCs. Taken together, we propose that the changes in Stat92E pairing status is an intrinsically programmed mechanism for enabling prompt cell fate switch during the differentiation of stem cells.

Fig. 1. The Stat92E locus shows a distinct pairing pattern after ACD.

a A schematic of the early germline differentiation in Drosophila testis. b A representative image of Stat92E nascent transcript and mRNA in GSC. Lower panels show magnified regions from insets in the top images. Nascent transcripts are detected by both an intron probe (red) and an exon probe (green) in the nucleus. c Representative images of the nascent transcript of Stat92E (magenta, white arrowheads) in GSC (Left) and GB (Right). d Violin plots show distances between Stat92E nascent transcript puncta in GSCs vs. GB-SGs. The p value was calculated with two-tailed Student’s t-test. e Estimated cell cycle stage of a GSC and GB pair by cell connectivity. f) A representative image of a GSC-GB pair in S-phase. White arrowheads indicate multiple puncta of nascent transcript (magenta). g A testis-tip image of Stat92E mRNA FISH. GSCs and CySCs are marked by white lines. h Examples of mRNA quantification from mid-plane of cells in indicated stages. mRNA molecules (smFISH dots) are encircled by solid-line circles. i Quantification of Stat92E mRNA (smFISH dots/plane) at the indicated stages. SIs (see texts) are shown in boxes above bars. j Representative images of Stat92E nascent transcript (magenta, white arrowheads) in the cells at the indicated stages. k Quantification of the intensity of Stat92E nascent transcript. Y axis indicates fluorescence intensity of nascent transcript signal in indicated stage of germ cells divided by CySCs’ (see Method for details). To plot intensity/single allele, measured intensities of paired cases were divided by two. Note that the data in all stages contain fractions with extremely low nascent transcript level, likely reflecting S-phase cells in which transcription is silent. In all RNA FISH experiments, nos > αTubulin-GFP flies were used to identify stages of germline (blue in b and g, green in c, f and j). Scale bars represent 2μm. Box plots show 25–75% (box), median (band inside) and minimum to maximum (whiskers) with all data points. The adjusted p values were calculated by one-way anova with Dunnett’s multiple comparisons for comparing each dataset with GSC data in (i) and (k). All plotted data points are provided in Source Data.

Fig. 2. The change in Stat92E pairing states is locus- and cell type-specific.

Left columns (a, d, g, j, m); A schematic of the Stat92E locus on chromosome3 (a). Estimated TAD boundaries and the region recognized by the Stat92E OligoPaint DNA FISH probe sets are shown at the top (a). A schematic of the lacO 98F6/99A7 insertion (d), the lacO 60 F insertion (g), the position of the Stat92E locus on chromosome 3 and on the TM3 balancer (j), and the Stat92E deficiency, Df(3 R)BSC516 (m). Middle columns (b, e, h, k, n); b, k, n Representative images of DNA FISH targeting the Stat92E locus in the indicated stages of germ cell development in the indicated genotypes; (b) wild type (yw), (k) heterozygous for TM3 (+/TM3), (n) heterozygous for Stat92E deficient allele (Df(3 R)BSC516/+), e, h Representative images of DNA FISH targeting lacO locus in the indicated stages of germ cells development in lacO 98F6/99A7 homozygous (e) or lacO 60 F homozygous (h) genotypes. In all DNA FISH samples, germ cells were visualized by Vasa staining (cyan). DNA FISH signals are shown in red (pointed by white arrowheads). Representative pairing states are shown in lower left corner of each image. All scale bars represent 2 μm. Violin plots (right columns, c, f, I, l, o); Violin plots show the distance between puncta of DNA FISH corresponding to the experiment shown in the middle panels (see details for Supplementary Fig. 2). Although most of the cells in Df(3 R)BSC516/+ flies showed only single spot plotted as distance zero (o), we also detected cells which had 2 puncta within a single nucleus in a low frequency (~10%), likely representing separated sister chromatids. Violin plots show KDE (kernel density estimate) and quantile lines and the width of each curve corresponds with the frequency of data points. The adjusted p values are calculated using Dunnett’s multiple comparisons comparing with GSC data for c. For other graphs, p-values were calculated by comparing with data shown in c using Šidák’s multiple comparisons. All plotted data points are provided in Source Data. Number of scored cells, which are randomly chosen from at least 10 testes for each experiments, is shown for each data point.

Fig. 3. The change in Stat92E pairing states does not reflect the difference in cell-cycle stages.

a A representative image of Stat92E nascent transcript (red) in Fly-Fucci testis. GFP-E2F1-230 (blue) is positive in G2-phase and negative in S-phase cells. White dotted lines encircle germ cells in indicated cell cycle stages. DAPI is shown in green. Right panel shows GFP channel. Stat92E nascent transcript was typically weak or undetectable in S-phase cells. b Representative images of G2 phase germ cells in indicated stages. Stat92E intron FISH signal (white arrowheads) indicate pairing states at the indicated stage of germ cells. Representative pairing states are shown in the upper right corner of each image. Lower panels show GFP channel. c Violin plots show distances between Stat92E nascent transcript puncta at the indicated stages of germ cell development measured in all cell-cycle stages or only in G2-phase cells. Violin plots show KDE and quantile lines and the width of each curve corresponds with the frequency of data points. The adjusted p values were calculated with Šidák’s multiple comparisons. All plotted data points are provided in Source Data. Number of scored cells, which are randomly chosen from at least 10 testes for each experiments, is shown for each data point. Scale bars in (a) represent 10 μm. Scale bars in b represent 2 μm.

Fig. 4. The change in Stat92E pairing states is required for subsequent silencing of transcription.

a A schematic of BacTg/Df(3 R)BSC516 (BacTg/Df) genotype. b Representative images of Stat92E DNA FISH puncta (red, white arrowheads) at the indicated stages of germ cells (Vasa, cyan) in BacTg/Df genotype. c Violin plots show distances between DNA FISH puncta. d Representative images of Stat92E mRNA (exon probe smFISH, green) at the indicated stages of germ cells in indicated genotypes. e Quantification of Stat92E mRNA levels. Y axis values are the number of mRNA dots present in a middle plane of the cell (dot#/plane). SIs (see text) are shown in boxes above bars. f Representative images of Stat92E nascent transcript (intron probe, red) at the indicated stages of germ cells in BacTg/Df testes. DAPI (cyan). g, h, i Representative images of Stat92E DNA FISH (red, white arrowheads) at the indicated stages of germ cells (Vasa, cyan) in indicated genotypes. j Violin plots show distances between DNA FISH puncta. k Quantification of Stat92E mRNA levels. Y axis values are the number of mRNA dots scored in a middle plane of the cell (dot#/plane). SIs (see text) are shown in boxes above bars. Adjusted p values are provided for comparison between control (wild type, yw) and BacTg/Df data in (c) and (e) and comparison between temperature-shifted experiment (temp shift) and no temperature-shift control (no temp shift) in (j) and (k). The adjusted p values were calculated using Šidák’s multiple comparisons. Violin plots show KDE and quantile lines and the width of each curve corresponds with the frequency of data points. Box plots show 25–75% (box), median (band inside) and minimum to maximum (whiskers) with all data points. All plotted data points are provided in Source Data. Number of scored cells, which are randomly chosen from at least 10 testes for each experiments, is shown for each data point. For RNAi experiments, temperature sensitive nosGal4 driver, nosGal4^ts, was used. All scale bars represent 2 μm. Representative pairing states are shown in lower left corner of each image.

Fig. 5. Stat92E nascent transcript level differs in paired and unpaired conditions.

a Schematic of cell cycle stages of a GSC and GB pair. b Representative images of Stat92E nascent transcripts in GBs. Left and right panel show paired or unpaired Stat92E homologous alleles, respectively. Red arrowheads indicate Stat92E nascent transcripts. c Quantification of nascent transcript intensity (measured as a ratio relative to nascent transcript intensity of CySC’s, see Method for details) in paired and unpaired conditions of GBs. Left two columns show intensity/punctum. Right two columns show intensity/allele. To plot intensity/allele, measured intensities of paired cases were divided by two. d Representative images of Stat92E DNA FISH in GBs. Left and right panel show paired or unpaired Stat92E homologous alleles, respectively. Red arrowheads indicate Stat92E locus visualized by DNA FISH. e Quantification of DNA FISH signal intensity in paired and unpaired conditions of GBs. Left two columns show intensity/punctum. Right two columns show intensity/allele. To plot intensity/allele, measured intensities of paired cases were divided by two. f A schematic of the intensity measurement results shown in (c) and (e). Box plots show 25–75% (box), median (band inside) and minimum to maximum (whiskers) with all data points. The p values were calculated by two-tailed Student’s t-tests. All scale bars represent 2 μm. Number of scored cells, which are randomly chosen from at least 10 testes for each experiments, is shown for each data point.

Fig. 6. Regulation of pairing requires the Stat92E enhancer but not transcription.

a A schematic of the Stat92E⁰⁶³⁴⁶ allele. b, c Representative patterns of Stat92E DNA FISH (red, white arrowheads) (b) or nascent transcript (c) in 2-cell SGs (cyan) in Stat92E⁰⁶³⁴⁶/+ flies. d Measured distances between puncta using DNA FISH or intron RNA FISH of Stat92E⁰⁶³⁴⁶/+. e Representative images of Stat92E DNA FISH (red, white arrowheads) in Stat92E⁰⁶³⁴⁶/+. Vasa (cyan). f Violin plots show distances between Stat92E DNA FISH puncta in Stat92E⁰⁶³⁴⁶/+. g Representative images of Stat92E mRNA FISH (green) in Stat92E⁰⁶³⁴⁶/+. nos > αTubulin-GFP (magenta). h Quantification of Stat92E mRNA levels in Stat92E⁰⁶³⁴⁶/+. Y axis values are the number of mRNA dots present in a middle plane of the cell (dot#/plane). SIs (see text) are shown in boxes above bars. i, l Experimental design of the flySAM (i) and the CRISPRi (l). j, m Representative images of Stat92E mRNA (green) in nos > flySAM (j) or nos > CRISPRi (m). nos > αTubulin-GFP (magenta). o Quantification of Stat92E mRNA levels in control (nos > αTub-GFP), nos > flySAM or nos > CRISPRi. Y axis values are the number of mRNA dots present in a middle plane of the cell (dot#/plane). SIs (see text) are shown in boxes above bars. k, n Representative images of Stat92E DNA FISH (red, white arrowheads). p Violin plots of distances between DNA FISH puncta. All scale bars represent 2 μm. Number of scored cells, which are randomly chosen from at least 10 testes for each experiments, is shown for each data point. Violin plots show KDE and quantile lines and the width of each curve corresponds with the frequency of data points. Box plots show 25–75% (box), median (band inside) and minimum to maximum (whiskers) with all data points. All plotted data points are provided in Source Data. Representative pairing states are shown in lower left corner of each image. The p values were calculated by two-tailed Student’s t-tests for d, and the adjusted p values were calculated using Šidák’s multiple comparisons for other graphs. All scale bars represent 2 μm.

Fig. 7. Stat92E pairing is under the control of asymmetric histone inheritance.

a, b Representative images of Stat92E DNA FISH (red, white arrowheads) at the indicated stages of germ cells from nos > H3-GFP (a) or nos > H3T3A-GFP (b). Vasa (cyan). c, d Violin plots of distances between Stat92E DNA FISH puncta (c) or between the lacO 60 F locus in lacO 60 F homozygous flies, both expressing nos > H3 or nos > H3T3A. e, f) Representative images of Stat92E mRNA (exon probe smFISH; green) at the indicated stages of germ cells from nos > H3-GFP (e) or nos > H3T3A-GFP (f). GFP (magenta). g Quantification of Stat92E mRNA levels in control (nos > H3-GFP) or nos > H3T3A-GFP expressing germ cells. Y axis values are the number of mRNA dots present in a middle plane of the cell (dot#/plane). SIs (see text) are shown in boxes above bars. h, i, j Representative images of Stat92E DNA FISH (red, pointed by white arrowheads) at the indicated stages of germ cells from control (h), haspin RNAi (i) or Cal1 RNAi (j) testes. Vasa (cyan). k Violin plots of distances between DNA FISH puncta. Temperature-sensitive nosGal4^ts driver was used for all RNAi experiments. Temperature shift was performed in 29 °C for three days for Cal1 RNAi to avoid germ cell loss phenotype, five days for haspin RNAi. P values were provided by comparison with no-temp-shift controls. l A schematic shows observed change of local pairing states of the Stat92E gene and effect on gene silencing. Violin plots show KDE and quantile lines and the width of each curve corresponds with the frequency of data points. Box plots show 25–75% (box), median (band inside) and minimum to maximum (whiskers) with all data points. All plotted data points are provided in Source Data. Number of scored cells, which are randomly chosen from at least 10 testes for each experiments, is shown for each data point. For all graphs, adjusted p values were calculated using Šidák’s multiple comparisons. All scale bars represent 2 μm. Representative pairing states are shown in lower left corner of each image.