Figure 1.
Follicle stem cells undergo neutral drift.
A. A Drosophila ovariole stained for Hu li tai shao (green) to highlight cell membranes, Traffic jam (red) to highlight somatic cell nuclei and DAPI (blue) to highlight DNA. In each ovariole, new follicles are produced by the germarium (boxed region, diagramed below) at the anterior tip. Follicles move toward the posterior as they mature. The germarium is divided into four sections as indicated, and the FSCs reside at the Region 2a/2b border. Anterior is to the left. B. Mosaic, uniformly marked (GFP-), and uniformly unmarked (GFP+) wildtype ovarioles stained for FasIII (red) to label follicle cells, GFP (green) to mark FSC clones, and DAPI (blue). C. Mosaic ovarioles become uniformly marked when the GFP+ FSC is replaced by a daughter of the GFP- FSC, or uniformly unmarked when the GFP- replaced by a daughter of the GFP+ FSC. These events are referred to as clone expansion (r+) or clone loss (r-), respectively. D. An alignment of the observed fraction of ovarioles with 0, 1 or 2 GFP- FSCs at 7, 14 and 21 dphs with the values predicted by the neutral competition model. The points indicate the actual data, and the error bars indicate the S.E. The solid lines indicate the values predicted by the model and the grey shaded areas indicate the 95% confidence interval ranges.
Table 1.
Quantification of marked FSC clone frequency at 7, 14 and 21 days post heat shock.
Figure 2.
FSCs have basolateral junction proteins and broadly distributed adherens junctions.
A. Diagram of the Drosophila germarium. Inset shows the localization of Lgl, Dlg and Scrib in follicle cells (white), which is similar in FSCs and downstream follicle cells. B–D. Wildtype germaria stained for Scrib (B), Lgl (C), or Dlg (D) (green) and DAPI (blue). The regions inside the dashed boxes in B′–D′ are magnified in B–D. E. Diagram summarizing the co-localization (yellow) of Arm (green) and DE-Cad in follicle cells. F–G. Germaria with a mature LacZ+ FSC clone (white) stained for Arm (green) and DE-Cad (red). F. Adherens junctions are distributed broadly in FSCs (white dotted line) and are restricted to the apical-lateral junctions in downstream follicle cells (orange triangles). The arm and DE-cad channels are shown separately with the LacZ channel in F′ and F″, respectively. G. A Notum-LacZ germarium stained for LacZ (red) to identify the FSC, DE-cad (green) and DAPI (blue). The boxed region in G is magnified in G′ and G″. Again, DE-cad is distributed in a broad streak along the anterior lateral surface of the FSC (white dotted line) and restricted to the apical-lateral junctions in downstream follicle cells (orange triangles). Anterior is to the left. Scale bar represents 5 µm.
Figure 3.
Bazooka is not localized to the apical domain of FSCs.
A. A germarium with a mature LacZ+ FSC clone (white) stained for Baz (green) and DAPI (blue). Baz is distributed diffusely in the FSC and localized to the apical domain in downstream follicle cells (orange triangles). B. A Notum-LacZ germarium stained for LacZ (red) to identify the FSC, baz (green) and DAPI (blue). Again, baz is diffuse and difficult to detect in the FSC, but upregulated in the downstream follicle cells (orange triangles). The boxed regions in A and B are magnified in A′ and A″, and B′ and B″ respectively. A″ and B″ show the Baz channel only. Germ cell cysts are outlined in solid white lines and stages are indicated. Anterior is to the left. Scale bar represents 5 µm.
Figure 4.
lgl and dlg mutant FSCs, but not scrib or baz mutant FSCs, are hypercompetitive.
A. The competition bias of marked (wildtype or mutant) FSC clones for each genotype relative to the unmarked (wildtype) FSCs, inferred by fitting the clone dynamics to the FSC competition model (Fig. 1C and Methods). A bias of 0 indicates that the marked FSCs undergo neutral competition with the unmarked FSCs; a positive bias indicates that the marked FSCs are more competitive than the unmarked FSCs; and a negative bias indicates that the marked FSCs are less competitive than the unmarked FSCs. Error bars = S.E. Stars indicate that the bias is significantly different from zero (p<0.001, Table S3). B. The rate of replacement of marked (wildtype or mutant) FSCs and unmarked (wildtype) FSCs for each genotype is similarly obtained from the model. The expansion/loss rates correspond respectively to r+, r- in Fig. 1C. Error bars = S.E. C. The observed and predicted fraction of ovarioles with 0, 1 or 2 GFP- FSCs at 7, 14 and 21 dphs for each genotype. Data points indicate experimental measurements (error bars = S.E.), faint lines and grey regions indicate the mean and 95% confidence intervals assuming no change in FSC behavior from wildtype, except for labeling efficiency. Solid lines show the Maximum Likelihood Fits with changes in both the competitive bias and the average FSC replacement rate according to the parameter values in panels A,B. The observed values for lgl1 and dlgm52 deviate substantially from the predicted values whereas the observed values for scrib2 and baz1 are closely aligned with the predicted values. The observed values for scrib1 deviate substantially from the predicted values because the replacement rate, but not the bias, is significantly different from wildtype.
Table 2.
Quantification of phosphohistone H3+ follicle cells in the germarium.
Figure 5.
Hypercompetitive prefollicle cells accumulate in the FSC niche region.
A–D. Germaria with mature GFP- wildtype (A), lgl1/1 (B), dlgm52/m52 (C), or scrib1/1 (D) FSC clones 14 days post heat shock stained for FasIII (red), GFP (green), and DAPI (blue). Wildtype clones typically contained only one GFP- prefollicle cell along the Region 2a/2b border (orange triangle, A) whereas lgl1/1 and dlgm52/m52 clones often contained 2 or more GFP- prefollicle cells in this region (orange triangles, B and C). scrib1/1 clones typically produced prefollicle cells that only migrated along the side of the germarium (white arrows) and rarely produced prefollicle cells that migrated along the Region 2a/2b border. Images of the full germarium is shown in the panel below each magnified region. Anterior is to the left. Scale bar represents 5 µm. E. Quantification of the average number of GFP- (wildtype or mutant) prefollicle cells along the Region 2a/2b border in mosaic germaria of the indicated genotypes. Error bars represent S.E. and significance was determined using a two-tailed Students T-test. * indicates p<0.05 and ** indicates p<0.01.
Figure 6.
Hypercompetitive prefollicle cells have reduced castor expression.
Germaria with mature wildtype (A), lgl1 (B), dlgm52 (C), or scrib1 (D) FSC clones stained for castor (red), GFP (green) and DAPI (blue). In wildtype and scrib1 clones, cas expression is low in FSCs (white arrows) and higher in prefollicle cells (orange arrowheads) whereas in lgl1 and dlgm52 clones, cas expression is low in both FSCs (white arrows) and prefollicle cells (orange arrowheads). The uniformly bright polar and stalk cells (orange arrows) within clones of all genotypes served as a control for antibody staining and exposure times across samples. In addition, the high levels of cas expression in lgl1 and dlgm52 polar and stalk cells, indicate that cas expression, though delayed, eventually reaches normal levels in these mutant clones.