(A and B) (A) Wild-type and (B) slbo mutant border cells (arrows) labeled with c522-GAL4-driven expression of UAS-actinEGFP (green); late stage-9 egg chambers. Anterior is oriented toward the left; border cells migrate toward the oocyte (right); phalloidin labels F-actin (red). The scale bars = 20 μm.
(C) Wild-type stage-9 egg chamber with follicle cells (arrowhead) labeled by c323-GAL4 + UAS-actinEGFP.
(D) Merged fluorescent and bright-field images of a GFP-labeled border cell cluster in the crude suspension of dissociated cells.
(E) FACS plot used to identify GFP-labeled border cells. GFP-positive events are 0.4% of the total population.
(F) Purified GFP-border cell clusters after FACS.
(G) FACS plot used to purify GFP-labeled follicle cells (~20% of the total population).
(H) Electrophoretic analysis on a few picograms of total RNA extracted from border cells or follicle cells. The ribosomal RNAs are visible as a double band.
(I) Electrophoretic analysis of amplified antisense RNA generated from 30 ng total RNA from each sample after amplification and biotin labeling.
(J–L) scatter plots of gene expression comparisons from different hybridizations. Represented by dots are the 14,010 probe sets present in the Drosophila Genome Array. (J) Technical replicate; two experiments were performed on one RNA collection from wild-type border cells. (K and L) Hybridization profiles of wild-type border cells (WT BCs) versus follicle cells (FCs) and wild-type versus slbo mutant border cells (slbo BCs), respectively.
(M) Validation for selected genes by quantitative real-time RT-PCR, done in triplicate on each of three independent RNA samples (for CG9629 and Hex-C, only on two samples) for each genotype. Data are reported as fold change between wild-type border cells (set to 1) and follicle cells or slbo mutant border cells. “*” indicates a transcript not detected.

(A) A Stathmin-specific antibody demonstrates high expression in migrating border cells (arrow) compared to follicle cells (arrowhead). (B–D) The Stathmin protein level (red) is severely reduced in slbo mutant border outer border cells (arrows), but it is unaffected by slbo in anterior polar cells (arrowheads). slbo mutant cells are marked by the absence of GFP (green). DAPI (blue): nuclei. In (C), cells are of the slbo mutant; in (D) and (E), some border cells are wild-type for comparison.
(E) The stathmin locus. Stathmin^ExC is a viable, fertile deletion of the first two exons of the stathmin C transcript; stathmin^L27 deletes the entire stathmin gene as well as CG31642, Arc-p20, CG9226, and CG9227.
(F) A stathmin^L27 mutant border cell clone, marked by the absence of GFP (arrow), does not migrate (n = 3).
(G) Normal migration in the egg chamber from a stathmin^exC/stathmin^L27 female, with complete loss of the stathmin C transcript. (H and I) Border cell-specific RNAi of stathmin decreases Stathmin protein levels in a sensitized background, stathmin^exC/stathmin^L27; UAS-stathminRNAi/slbo-GAL4 in (I); compare to wild-type (H). In (B)–(D) and (H)–(I), right panels shows anti-stathmin alone, and the left panels show a merged image.
(J and K) Quantification of migration at (J) stage 9 and (K) stage 10 in border cells with reduced Stathmin levels (stathmin^exC/stathmin^L27; UAS-stathminRNAi/slbo-GAL4) compared to control (stathmin^exC/stathmin^L27; TM3/slbo-GAL4) and “rescue” (UAS-stathmin/+; stathmin^exC/stathmin^L27; UAS-stathminRNAi/slbo-GAL4).
(L) Overexpression of stathmin in slbo mutant border cells does not rescue migration. Quantification of slbo¹³¹⁰,slbo-GAL4/slbo¹³¹⁰ and slbo¹³¹⁰,slbo-GAL4/slbo¹³¹⁰;UAS-Stathmin egg chambers at stage 10.

(A and B) (A) Tropomyosin2-GFP (protein trap) in migrating border cells (arrow) compared to follicle cells (arrowhead) and (B) in the muscle sheath that surrounds egg chambers, counterstained for F-actin (phalloidin, red). The scale bars are 20 μm.
(C) Stage-10 egg chamber with an mhc¹ mutant border cell clone (marked by the absence of GFP) that shows normal migration.
(D) Stage-10 egg chamber with an mhc³ mutant border cell clone (marked by the absence of GFP) that is delayed (<50%).
(E) Late stage-9 egg chamber with a rols^T627 mutant border cell clone that shows defective migration (50%).
(F) Quantification of border cell migration in egg chambers at stage 10 for mhc¹, mhc³, mlc2^E38, up¹⁰¹ and rols^T627 mutant border cell clones, or in tm2³ homozygous mutant females. Controls for mhc¹, mhc³, and rols^T627 mutant clones were: mhc¹/+ (n = 46), mhc³/+ (n = 21), and rols^T627/+ (n = 20).

(A) six4 expression detected by RNA in situ hybridization.
(B) Eya expression detected by antibody staining. Early stages are to the left.
(C) Six4²⁸⁹ mutant follicle cell clone (marked by the absence of GFP, arrows) in the anterior of a stage-9 egg chamber stained with anti-Slbo (border cells, red in [C]) and anti-Fas3 (polar cells, blue in [C′]). Wild-type cells directly in contact with the mutant clone express Slbo (arrowheads).
(D) Stage-10 egg chamber with two border cell clusters (arrowhead: wild-type cluster, arrow: ectopically induced cluster).
(E and F) six4²⁸⁹ mutant follicle cell clones stained with anti-Eya (red). (E) Anterior tip and (F) posterior tip of egg chambers.
(G) Delayed migration in a stage 9 slbo¹³¹⁰, slboGAL4/UAS-HA-six4 egg chamber. HA-Six4 is nuclear.
(H) Quantification of migration in stage-10 egg chambers of the genotype slbo¹³¹⁰, slboGAL4/UAS-HA-six4 and control slbo¹³¹⁰, slboGAL4/+.

(A–C) Diagrams representing genes with significant and at least 2-fold change in RNA level in wild-type border cells (wt BCs) compared to follicle cells (FCs) and slbo mutant border cells (slbo BCs), respectively. (A) Genes upregulated in wild-type border cells. (B) Genes downregulated in wild-type border cells. (C) Overlap of genes expressed at a high level in wild-type border cells (blue) and genes expressed at a higher level in slbo mutant border cells (yellow).
(D–F) Pie charts showing the relative distribution according to seven functional categories of genes found upregulated at least 2-fold in wild-type border cells; the cytoskeleton-associated genes upregulated in both comparisons are listed
(G) Representation of the functional groups in the Drosophila genome. For (D)–(G), only genes with some functional annotation were included: D:216, E:66, F:118, and G:6537 genes.

(A–C) Egg chambers stained with antibody to Singed. (A) High Singed expression in wild-type stage-9 border cells (arrow) compared to follicle cells (arrowhead). (B and C) Merged and (B′ and C′) single channel images of a mosaic border cell cluster ([C] at high magnification); slbo^8ex2 mutant cells are marked by the absence of GFP and by a yellow arrow; wild-type cells are marked by white arrow. The scale bars are 20 μm.
(D) Stage-9 egg chamber stained with anti-Gelsolin. High Gelsolin expression is detected in polar cells within the border cell cluster (arrow) compared to follicle cells (arrowhead).
(E and F) Late stage-9 egg chambers stained with anti-Quail. (E) Wild-type. (F) Egg chamber with a quail germline mutant clone (GFP negative); Quail is detected in border cells (arrow).
(G) Quantification of border cell migration in egg chambers at stage 10, when migration is complete in wild-type.

(A–F) Quantification of migration at stage 10 for (A) aop (yan)^PB104 and (E) vrille² mutant border cell clones. Controls are aop^PB/+ and vrille²/+. (B) Stage-10 egg chamber with an aop^PB104 mutant border cell clone (arrow): delayed migration. (C) Late stage-9 egg chamber with an aop¹ mutant border cell clone. Mutant cells are marked by the absence of GFP; one cell is GFP positive. (D and D′)An aop¹ mutant follicle cell clone in a stage-10 egg chamber. Phalloidin (red) and DAPI (blue). The scale bar is 20 μm. (F) A vrille² mutant border cell clone marked by the absence of GFP (arrow).