Mosaic Analysis in the Drosophila Ovary Reveals a Common Hedgehog-Inducible Precursor Stage for Stalk and Polar Cells

Corresponding author: Hannele Ruohola-Baker, Department of Biochemistry, Box 357350, University of Washington, HSB J-581, 1959 Pacific St., Seattle, WA 98195., hannele{at}u.washington.edu (E-mail)

Communicating editor: T. SCHÜPBACH

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The fates of two small subgroups of the ovarian follicle cells appear to be linked: mutations in Notch, Delta, fs(1)Yb, or hedgehog cause simultaneous defects in the specification of stalk cells and polar cells. Both of these subgroups are determined in the germarium, and both cease division early in oogenesis. To test the possibility that these subgroups are related by lineage, we generated dominantly marked mitotic clones in ovaries. Small, restricted clones in stalk cells and polar cells were found adjacent to each other at a frequency much too high to be explained by independent induction. We therefore propose a model in which stalk cells and polar cells are derived from a precursor population that is distinct from the precursors for other follicle cells. We support and extend this model by characterization of mutants that affect stalk and polar cell formation. We find that ectopic expression of Hedgehog can induce both polar and stalk cell fate, presumably by acting on the precursor stage. In contrast, we find that stall affects neither the induction of the precursors nor the decision between the stalk cell and polar cell fate but, rather, some later differentiation step of stalk cells. In addition, we show that ectopic polar and stalk cells disturb the anterior-posterior polarity of the underlying oocyte.

THE differentiation of specialized cell types can be governed both by external signals and intrinsic differences resulting from asymmetric cell divisions. Unraveling how the complex interplay of extrinsic and intrinsic factors determines cell fates is a challenging problem in developmental biology (for a recent review see JAN and JAN 1998 ). In some tissues, for example, the Drosophila ommatidia, where the cells are not related by lineage, cell fates are determined entirely by external cues in the form of cell-cell signaling pathways (BECKER 1957 ; WOLFF and READY 1993 ). In contrast, yeast mother-daughter cell differentiation does not rely on external stimuli, but it is attributable to asymmetric division that leads to asymmetric partitioning of the Ash1 protein and, therefore, asymmetric expression of HO endonuclease (BOBOLA et al. 1996 ; SIL and HERSKOWITZ 1996 ). In some cases, such as Drosophila neuroblasts, which are lineage related, both internal and external cues are used (reviewed in GHO and SCHWEISGUTH 1998 ; JAN and JAN 1998 ). Asymmetric segregation of Numb protein during cell division leads to inactivation of Notch receptor in one of the daughter cells.

In the Drosophila ovary, the germline is surrounded by a somatic follicle cell layer, and signaling from this layer plays a key role in oocyte development. In particular, patterning of the follicle cell layer generates positional information that is transferred to the developing oocyte via signaling pathways, and that is instrumental for establishment of embryonic polarity. Therefore, it is pivotal to understand the patterning of the follicle cell layer. Early in oogenesis, the dividing somatic cells are largely undifferentiated and provide an excellent system in which to study cell fate decisions. It has been shown that cell-cell communication plays a major role in follicle cell differentiation (reviewed in RAY and SCHUPBACH 1996 ; RONGO and LEHMANN 1996 ). Mosaic analysis has shown that most specialized subgroups of follicle cells are not restricted by lineage (MARGOLIS and SPRADLING 1995 ).

Two small subgroups of follicle cells have been central to several genetic investigations: the polar cells and the stalk cells. Polar cells are sets of follicle cells located at the anterior and posterior tips of an egg chamber; stalk cells are a linear group of follicle cells that separate the egg chambers (Figure 1). Differentiation of these subgroups occurs when follicle cell intercalation separates egg chambers from the germarium, where germ line division takes place. Mutations in hedgehog (hh), Notch (N), Delta (Dl), or fs(1)Yb cause defects in the specification of stalk and polar cell number (RUOHOLA et al. 1991 ; XU et al. 1992 ; BENDER et al. 1993 ; JOHNSON et al. 1995 ; FORBES et al. 1996A , FORBES et al. 1996B ; LARKIN et al. 1996 ). daughterless (da) and toucan (toc) are also involved in this process; however, their effect on both stalk and polar cell fates remains to be characterized (CUMMINGS and CRONMILLER 1994 ; GRAMMONT et al. 1997 ). Models for differentiation of the stalk and polar cell subgroups have been generated on the basis of genetic data (RUOHOLA et al. 1991 ; JOHNSON et al. 1995 ; LARKIN et al. 1996 ). These models imply a lineage relationship between the two subgroups and suggest that some of the asymmetric differentiation cues might be intrinsic. However, the lineage relationship between these subgroups has not been tested.

View larger version (39K): In this window In a new window Download PPT slide   Figure 1. (A) The system used to mark mitotic clones (adapted from MARGOLIS and SPRADLING 1995 ). The lacZ gene is constitutively driven by the tubulin promoter because of an Flp-induced recombination event. The daughter cell that inherits the recombinant construct is dominantly marked, as are its progeny. (B) The time scale for different developmental steps in oogenesis. tf, terminal filament cells; gsc, germline stem cells; fc, follicle cells; gc, germline cells; sc, stalk cells; pc, polar cells; S6 and S8, stages 6 and 8 of oogenesis. Unit is a PC-SC-PC unit on each side of a stage 6 oocyte.

Stalk cells and polar cells are determined in the germarium. Previous clonal analysis showed that polar cells stop dividing in the germarium long before the rest of the follicle cells surrounding the egg chamber (MARGOLIS and SPRADLING 1995 ). Here we confirm this result and further show that stalk cells cease division early in oogenesis. To test the possibility that stalk and polar cells are related by lineage, we generated dominantly marked mitotic clones in ovaries. Small, restricted clones in stalk cells and polar cells were found adjacent to each other at a frequency much too high to be explained by independent induction. We, therefore, propose a model in which stalk cells and polar cells are derived from a precursor population that is distinct from precursors for other follicle cells. We support and extend this model by characterization of mutants that affect stalk and polar cell formation.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Fly stocks:
All lines were maintained at room temperature on standard cornmeal and molasses agar medium with dry yeast added. The following mutant alleles of stall were used in this study: WU40, AWK26, PA49, and PH57 (SCHUPBACH and WIESCHAUS 1991 ). The aTub84BFRTy+FRThh- line was a generous gift from G. Struhl. The enhancer trap lines 10613 and 10475 mark inner sheath cells and are insertions in the 5' end of the patched gene (MARGOLIS and SPRADLING 1995 ; T. MAHOWALD, personal communication); A101 marks polar cells. 93F and l(3)1344 mark stalk cells (RUOHOLA et al. 1991 ; FORBES et al. 1996A ; LARKIN et al. 1996 ; ZAFFRAN et al. 1997 ). GAl4(127) marks terminal filament cells when used to drive the UAS-lacZ reporter gene. Kin:ßgal line KL-32 (CLARK et al. 1994 ) was used. The lines y-w-; x-15-33, y+/Cyo, y-w-; x-15-29, w+/Cyo and hsFLP, Mkrs/nkd were generated by Harrison (HARRISON and PERRIMON 1993 ) and were a gift from J. Margolis. x-15-33 and x-15-29 are the lines that carry the FRT sites, arranged as depicted in Figure 1. These lines were crossed to generate y-w-; x-15-33, y+/x-15-29, w⁺; hsFLP, MKRS/+ flies for clone induction.

Heat shocks:
To induce mitotic recombination, 1- to 4-day-old adult y-w-; x-15-33, y+/x-15-29, w⁺; hsFLP, MKRS/+ females were placed on wet yeast for 1 day and then subjected to a 1-hr heat shock at 37° in a circulating water bath. Afterward, the flies were maintained at 25° and turned over onto fresh yeast daily. Flies were allowed to develop for 2 days and were then dissected in 1x PBS.

A previous study using this system in follicle cells (MARGOLIS and SPRADLING 1995 ) found that clone induction was limited to a short time after heat shock. Additionally, the heat shock itself did not cause noticeable developmental arrest in ovaries, and the labeling frequency was approximately constant throughout oogenesis. Thus, this Flp/Frt system has proven to be a well-behaved tool for the clonal analysis of follicle cells. Furthermore, ß-galactosidase expression in recombinant clones is suitable for the analysis of stalk and polar cell subgroups.

To induce Hedgehog expression, HS-FLP; Tub84BFRTy+FRThh females were heat shocked for 15 min at 39° or 30 min at 37°. Flpase activity places the hh gene downstream of the Tub promoter, permitting persistent, ectopic expression of hh. Flies were allowed to develop 24, 48, or 72 hr at room temperature before dissection. Receipt of the Hh signal was followed by patched (ptc) enhancer trap lines.

Staining and visualization:
Immunocytochemistry, X-gal stainings, and light and confocal microscopy were carried out as described in LARKIN et al. 1996 , with the following exception: after x-gal staining, ovaries were postfixed for 15 min in 2% glutaraldehyde in 1x PBS.

For chromosome visualization by 4',6-diamidino-2-phenylindole (DAPI) staining, ovaries were dissected and fixed for 1 hr in 4% paraformaldehyde. After rinsing in PBS, ovaries were incubated in 0.01 mg/ml DAPI in PBS for 2 min. They were then rinsed and mounted on 70% glycerol slides. Propidium iodide (PI) chromosome staining was performed in the same manner, with the addition of a 30-min incubation in 0.5 mg/ml RNAse (Sigma, St. Louis) in PBS, followed by a 10-min incubation in 1 mg/ml PI.

Calculations:
The units counted in these experiments cover the posterior end of the earlier egg chamber, connecting stalk cells and the anterior end of the older egg chamber [(polar cells-stalk cells-polar cells (PC-SC-PC)]. Each stage 6 egg chamber contains two units. An alternative unit could be considered as the stalk and polar cell subgroups on opposite sides of the stage 6 egg chamber (SC-PC-(egg chamber)-PC-SC). However, the former definition of a unit is supported by the fact that coincident clones occurred 24 times in our experiments. In contrast, only five coincident clones occurred in sc or pc subgroups on opposite sides of stage 6 egg chambers in the same experiments.

We used the chi-square test in our analysis of stalk and polar cell correlation (for review see TAYLOR 1997 ). This analysis approximates how observed data relate to the expected outcome of the experiment. In the mosaic clone experiment, the observed data are the number of times stalk and polar cell clones occur separately vs. together; the expected outcome is predicted from independent induction of clones (see Table 2 and Table 3). Because the chi-square is an approximate test, we verified the results of the chi-square test using Fischer's exact test. We obtained a value of <0.001 for the former and 1 x 10^-16 for the latter, indicating the P value obtained by both methods is highly significant.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Stalk cells and polar cells cease dividing in the germarium:
The Drosophila ovary is composed of ~15 ovarioles. Each ovariole consists of two major parts, the germarium and the vitellarium. The germarium is located at the anterior tip of the ovariole, and it contains stem cells for both germline and somatic cells. A germline stem cell divides to produce a daughter cell, which undergoes four rounds of division in region 1 of the germarium to form a 16-cell germline cyst, which is enveloped in region 2b (see Figure 1B) by a layer of somatic follicle cells. An egg chamber is formed when somatic cells anterior to the cyst in region 3 of the germarium intercalate to form a mature stalk consisting of five to seven specialized stalk cells (reviewed by KING 1970 ; SPRADLING 1993 ) that separate adjacent egg chambers. Polar cells are located at the anterior and posterior poles of egg chambers and have a round morphology distinct from that of the columnar follicle cells (BROWER et al. 1981 ; RUOHOLA-BAKER et al. 1994 ). The statement "early oogenesis" in this study includes stages from the germarium to stage 6 of oogenesis.

Previous work has demonstrated by clonal analysis that polar cells cease division in the germarium (MARGOLIS and SPRADLING 1995 ). No such direct evidence has been shown for stalk cell divisions, although stalk cell morphology strongly suggests that these cells cease division early in oogenesis. To address this question, we used the FLP/FRT system (GOLIC 1991 ) modified by HARRISON and PERRIMON 1993 to induce clones in mitotic cells. This system uses a heat shock promoter to drive transcription of the Flpase protein, which catalyzes recombination specifically between FRT sites in mitotic chromosomes. Recombination places the lacZ gene downstream of an 84BTubulin promoter, thus dominantly and persistently marking the target cell and its progeny (Figure 1A). This system has previously been shown to exhibit no background recombination in the absence of heat shock, and to induce mitotic recombination only in actively cycling cells (HARRISON and PERRIMON 1993 ; MARGOLIS and SPRADLING 1995 ).

We analyzed the behavior of stalk cells and polar cells by examining ovaries dissected 2 days after heat shock and scoring clones in stage 6 egg chambers and their flanking stalks. We chose stage 6 because it is the most mature stage around which stalks are frequently visible. According to previously reported timing of oogenic stages (KING 1970 ; SPRADLING 1993 ), the progenitors of the clones we scored were in region 2b of the germarium (Figure 1B) at the time of FLP-induced recombination. Because approximately five rounds of division occur between region 2b and the end of stage 6 (MARGOLIS and SPRADLING 1995 ), we expected to observe follicle cell clones of ~16 cells in size. Significantly smaller clones should only appear in groups of follicle cells that cease dividing in the early stages of oogenesis. In keeping with our expectations, the average clone size in stage 6 follicles was 14.1 cells (Table 1). Smaller clones (approximately two cells) were primarily restricted to both putative polar and stalk cells (Table 2; Figure 2D), indicating that both of these groups cease division during or shortly after region 2b of the germarium, the calculated time of heat shock.

View larger version (88K): In this window In a new window Download PPT slide   Figure 2. Stalk cells and polar cells cease dividing early. Small clones (on average two cells) were detected in polar cells (A) and stalk cells (B and C) at a much higher frequency than in the main body follicle cells (D), suggesting that sc and pc subgroups ceased division early in oogenesis. The few extremely large clones (>60) detected in the main body follicle cells may result from multiple clones existing in close proximity. We also analyzed the frequency of small clones in the main body follicle cells as an indication of the amount of residual Flp activity after heat shock. If the extremely small clones in stalk and polar cell groups were attributable to residual Flpase, they should have occurred with approximately the same frequency, on a per cell basis, as small clones in the main body follicle cells. However, the frequency of small clones in main body follicle cells was 124 times lower than the frequency detected in the stalk cells, suggesting that the small clones detected in the stalks are due to early cessation of cell division and not to late induction via Flpase activity.

Stalk cells stop dividing late in the germarium:
Because stalks between stage 5 and 6 egg chambers have the same number of cells as those between stage 1 and 3 chambers, it is thought that stalk cells do not divide after germarium. To experimentally refine the window of time in which stalk cells stop dividing, both DAPI and PI chromosome stains were used to visualize mitosis. Metaphase and anaphase chromosomes are highly condensed and appear significantly brighter than interphase chromosomes in the presence of these dyes. As expected, such condensed chromosomes were clearly visible in ~1.2% (67/5700) of the division-competent follicle cells surrounding stage 1 and 2 egg chambers (data not shown). In contrast, we scored 1050 DAPI-stained mature stalk cells between stage 1 and 3 egg chambers without observing any evidence of mitosis. We therefore conclude that, unlike most of the follicle cells, stalk cells cease dividing in the germarium.

To identify the latest stage in the germarium in which divisions of stalk and polar cell precursors are detected, we marked these precursors with Big Brain (Bib) antibody (LARKIN et al. 1996 ; DOHERTY et al. 1997 ). Bib antibody specifically marks stalk and polar cell precursors in the germarium in a characteristic subcellular distribution pattern (LARKIN et al. 1996 ; see Figure 5I). We found condensed PI staining indicative of mitosis in Bib-positive cells in the early stages of egg chamber pinching (Figure 3), but not in later stages of pinching. We therefore conclude that the precursors for the stalk cells and polar cells are still competent to divide in the beginning of the egg chamber pinch.

View larger version (30K): In this window In a new window Download PPT slide   Figure 3. Bib-positive precursors still divide in the border of regions 2b and 3 in germarium. Wild-type ovaries are double stained with Bib antibody (A and D; red in C and F) and PI (B and E; green in C and F). Condensed chromatin (B and E), indicative of mitotic cells, is detected in precursor cells that align between the egg chambers before intercalation (C and F, arrows).

View larger version (64K): In this window In a new window Download PPT slide   Figure 4. Restricted clones in stalk cells and polar cells coincide. Small clones (schematic representation, A and B) that span stalk cells and polar cells can be continuous (A, C, E, and G, 70%) or discontinuous (B, D, F, and H, 30%). By stage 6 in oogenesis, the point of attachment of the stalk to the polar cells is released and the egg chamber can rotate (see Figure 1, S8). (E and F) Higher magnification of stalk cells. (G and H) Higher magnification of polar cells.

View larger version (73K): In this window In a new window Download PPT slide   Figure 5. Ectopic expression of Hedgehog in the germarium prolongs the Bib-positive precursor stage. Ovaries from wild-type (A, C, E, G, and I) and HSFlp; aTub84BFRTy+FRThh (48 hr after hh induction B, D, F, H, and J) animals stained with patched enhancer trap line (A and B), polar cell marker A101 (C and D), stalk cell marker 93F (E and F), and antibodies for FasIII (G and H) and Bib (I and J). The patched enhancer trap line (10613), which in wild-type ovarioles is expressed in the inner sheath cells (A, bracket), marks the cells that are affected by ectopic expression of Hh (B, bracket). Because of persistent ectopic expression of Hh, long, stalk-like structures are formed. The cells in these stalk-like structures do not differentiate to form the A101-positive polar cells (D, bracket) or 93F-positive stalk cells (F, bracket). Instead, the Bib- and FasIII-positive, morphologically distinct precursor stage is prolonged in the cells expressing ectopic Hh (H and J). In wild-type ovaries, Bib stains the apical tips of precursors for stalk and polar cells when they are aligned to intercalate in the germarium (horizontal pattern, I-I) and in the membranes between the cells in mature stalks (vertical pattern, I-II). Under sustained ectopic expression of Hh, apical precursor staining of Bib antibody staining is detected in follicle cells located between stage 1 to 4 egg chambers (J, brackets).

Restricted clones in stalk cells and polar cells coincide:
The fact that stalk and polar cells both finish dividing early allowed us to identify these two subgroups as small clones in our clonal analysis, and to look for a common lineage relationship between them. If such a lineage relationship exists, and if cells belonging to the common progenitor cells are dividing during stage 2b, then such cells should sometimes be marked in our mosaic experiments. A precursor thus marked should occasionally divide to produce both a marked stalk cell and a marked polar cell, leading to coincident restricted clones representing both cell types in PC-SC-PC units on either the anterior or posterior end of stage 6 egg chambers (Figure 1; see MATERIALS AND METHODS). A low number of coincident clones would also be expected as a result of independent mitotic events. We observed coincident clones (Figure 4) at a significantly high frequency (Figure 2A): the probability of the observed coincident stalk cell and polar cell clones arising by independent mitotic events is calculated to be <0.001 (Table 2; see MATERIALS AND METHODS). In contrast, the frequency of stalk cell clone and main body clone coincidence is far more consistent with independent induction (P ~ 1; Table 3). We conclude, therefore, that stalk cells and polar cells have a common precursor that is distinct from the precursors of main body follicle cells (see Figure 8).

View larger version (71K): In this window In a new window Download PPT slide   Figure 6. Hedgehog can induce ectopic stalk cells and polar cells that coincide with a defective oocyte anterior-posterior axis. Wild-type ptclacZ expression (A) is expanded to stage 1 to 8 egg chambers 3 days after Hh induction (B–D). In this time point, some of the cells located in long, stalk-like structures begin to express the stalk cell marker 93F (E, arrow, ~20 cells). Receipt of the ectopic Hh signal indicated by ectopic patchy expression of ptclacZ (B–D) induces ectopic polar (A101, F) and stalk cell (93F, G–I) fates in the follicle cell layer surrounding the germline. The anterior-posterior axis defect is detected in the underlying oocyte, as indicated by mislocalization of the Kin:ßgal fusion protein (K, 46% = 25/54 stage 8 to 9 egg chambers) and nucleus (E, arrow). Controls show normal posterior localization of Kin:ßgal (J). H is an enlargement of ectopic stalk cells (boxed area in I). F and G are 1/2, J and K are 1/4, A and B are 1/8, and H is 3x magnification of C–E and I.

View larger version (107K): In this window In a new window Download PPT slide   Figure 7. stall affects stalk cell migration but not the primary cell fate. DAPI staining of stall mutant (PH57/PA49) shows the accumulation of germline cysts in the enlarged germarium (A). The stall mutant (PH57/AWK26), which lacks morphological stalks, still allows polar cells and stalk cells to differentiate on the basis of A101 (B) and 93F (C, arrow) expression. All allele combinations of stall we analyzed expressed 93F and A101 in the enlarged germaria. 93F is also expressed in terminal filament (tf) cells (C, arrowhead). GAL4(127) is only expressed in the tf cells in wild-type and stall mutant ovarioles (D and E, arrowhead), indicating that no expansion of terminal filament cell fate is detected in the stall mutant. In all cases, anterior is on the left.

View larger version (16K): In this window In a new window Download PPT slide   Figure 8. Model for stalk cell and polar cell differentiation. Our data suggest that stalk cells and polar cells are related by lineage. Ectopic Hh and defects in fs(1)Yb and the Notch pathway affect the lineage by altering the precursor stage. daughterless and toucan also act in this process. In contrast, stall does not affect the lineage decision, but it is required later for stalk formation.

The restricted clones in stalk cells and polar cells that coincide in a unit (Figure 1) were either continuous (70%) or discontinuous (30%). Discontinuous clones could be generated either by stalk or precursor cell intercalation. All the coincident clones (n = 24) contained approximately two stalk cells and two polar cells, suggesting that after the cell fate decisions, polar cells and stalk cells undergo equal rounds of mitosis. Among the 24 coincident clones (Table 2), we never detected marked polar cells at both sides of a unit. These data suggest that each unit is generated by more than one precursor.

An earlier clonal analysis in follicle cells found that clone borders were inviolate: unmarked cells were never found in the middle of marked clones (MARGOLIS and SPRADLING 1995 ). Our observations of clones on the surface of the egg chamber further confirm this result. However, stalk cells migrate and intercalate to form the stalk, thus interrupting a clone by intercalating marked cells with unmarked cells. Moreover, while stalk cells always abut polar cells in early stages, this connection is released in later stages (see Figure 1B, stage 8), so it is not surprising that coincident clonal patches in stalk cells and polar cells are not always contiguous (Figure 4). Therefore, clonal borders can be violated in these subgroups.

Genetic mechanisms of stalk and polar cell differentiation:
A group of mutants that share a morphological defect in stalk formation have been identified (RUOHOLA et al. 1991 ; XU et al. 1992 ; BENDER et al. 1993 ; CUMMINGS and CRONMILLER 1994 ; JOHNSON et al. 1995 ; FORBES et al. 1996A ; LARKIN et al. 1996 ; GRAMMONT et al. 1997 ; LARKIN et al. 1999 ). Loss of function of Notch, Delta, fs(1)Yb, daughterless, hedgehog, toucan, or stall results in large germaria that accumulate cysts and do not form stalks; overexpression of constitutively active Notch (caN), Delta, toucan, or hh results in long, stalk-like structures. Having now shown that stalk cells and polar cells share a distinct precursor, we can ask if the lineage decision between these two groups is affected in each of the mutants bearing a morphological defect in stalk formation. We have previously shown that lack of Notch activity results in an excess of polar cells and loss of stalk cells because of a defect in a precursor stage (RUOHOLA et al. 1991 ; LARKIN et al. 1996 ).

Ectopic expression of Hedgehog in the germarium prolongs the precursor stage for stalk cells and polar cells:
The secreted protein Hedgehog (Hh) plays a role in regulating growth and patterning in a number of developing systems (for reviews see PERRIMON 1995 ; INGHAM 1995 ; JOHNSON and TABIN 1995 ). In oogenesis, hypomorphic alleles of hh produce large germaria that accumulate germline cells (FORBES et al. 1996A ). Transient ectopic expression of hh results in "long, stalk-like" structures and an excess of polar cells in ectopic positions (FORBES et al. 1996A , FORBES et al. 1996B ). We analyzed the cell fates in these situations in more detail using a persistent expression system (BASLER and STRUHL 1994 ; Figure 5 and Figure 6). Receipt of the Hh signal was followed by a ptc enhancer trap line 10613 (Figure 5A). Two days after Hh induction, long, stalk-like structures expressing the ptc marker were detected (Figure 5B). To analyze the identity of the cells in these long, stalk-like structures, we stained them with Big Brain antibody. Bib marks the precursors for stalk cells and polar cells, but not the cells that surround the rest of the egg chamber. The long, stalk-like structures generated by overproduction of Hh do not contain differentiated polar cells or stalk cells (FORBES et al. 1996A , FORBES et al. 1996B ; Figure 5D and Figure F). Instead, the follicle cells between 16 cell cysts resemble wild-type precursor cells for these populations in morphology and expression of Bib and FasIII proteins (Figure 5H and Figure J). However, not all of these cells remain in a precursor stage: at later time points, some of the cells in the long stalks "leak through" the Hh-induced precursor block and differentiate to form extra stalk or polar cells, further evidence that these cells are precursors for polar and stalk cells (Figure 6E, arrow).

Hedgehog can induce ectopic stalk cells and polar cells that coincide with a defective oocyte anterior-posterior axis:
Three days after persistent Hh induction, patchy patched expression, indicative of Hh action, extended to later-stage egg chambers (Figure 6, B–D). Interestingly, in addition to previously observed ectopic polar cells (FORBES et al. 1996A , FORBES et al. 1996B ; Figure 6F), we also observed ectopic stalk cells (Figure 6, F–I) in these egg chambers, suggesting that ectopic Hh can induce both fates. Therefore, ectopic Hh affects not only the stalk cell and polar cell fates in their normal location, but it can also induce these fates in ectopic locations, probably by transiently inducing the stalk and polar cell precursor stage. We have shown by clonal analysis that stalk and polar cells have a common precursor that is separate from a precursor for egg chamber follicle cells. These data show that ectopic Hh has a capacity to induce this precursor fate.

Coinciding with the ectopic polar cells, an anterior-posterior axis defect was detected in the underlying oocyte. The typical migration of the oocyte nucleus from the posterior to a dorsal-anterior location failed to occur (Figure 6F). In addition, Kin:ßgal fusion protein failed to localize posteriorly, and the oocyte microtubule network was defective (Figure 6K; data not shown). These phenotypes have been previously observed in mutants that compromise the epidermal growth factor receptor or Notch pathways in posterior follicle cells, thereby altering the follicle cell-oocyte signaling that is required for proper anterior-posterior axis formation (reviewed in ANDERSON 1995 ).

Stall affects stalk cell migration but not the primary cell fate:
Mutants in stall display defects in egg chamber separation from the germarium that are morphologically comparable to the Notch loss-of-function defect (SCHUPBACH and WIESCHAUS 1991 ; Figure 7A). We asked whether this defect was caused by the elimination of stalk cell fate, as in N^ts mutants (RUOHOLA et al. 1991 ). The expression of the enhancer trap line 93F (RUOHOLA et al. 1991 ; ZAFFRAN et al. 1997 ), which marks stalk cells, was examined in stall mutants. Although no stalk structures were detected, groups of 93F-positive cells were observed in the mutant germaria, suggesting that cells continue to acquire stalk cell fate (Figure 7C). The ratio of 93F-positive cells to germline cysts found in stall germaria is consistent with the number of cells found in wild-type stalks (data not shown). Because 93F marks the terminal filament cells as well as the stalk cells in wild-type ovaries, it was formally possible that the 93F-positive cells in stall germaria represented terminal filament cells. To rule out this possibility, we examined stall mutant ovaries with an independent terminal filament marker, Gal4(127) (Figure 7D). ß-Galactosidase protein visualization of stall mutant germaria carrying UAS-LacZ under control of Gal4(127) displayed only the wild-type terminal filament expression pattern (Figure 7E, arrowhead). In addition, some cells that expressed an exclusive stalk cell marker, l(3)1344, were detected in stall mutant germaria (FORBES et al. 1996A ; data not shown). We conclude that the 93F-positive cells in stall mutant germaria are stalk cells. Furthermore, no obvious defect in the polar cell population was detected in stall mutant germaria upon examining the expression of the A101 enhancer trap line (Figure 5B). These data suggest that the early lineage decision between polar cell and stalk cell fate occurs normally in the stall mutant. We conclude that stalk cells are defective in a subsequent differentiation or migration step.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We used clonal analysis to better characterize stalk and polar cell differentiation. We determined that stalk cells stop dividing during the early egg chamber pinch, when the associated cyst is in region 3 of the germarium. Furthermore, based on the tight correlation of restricted clones in stalk and polar cells, we propose a model in which these two groups share a common lineage (Figure 8). Supporting evidence for this hypothesis was obtained by analyzing the cell fate changes generated by ectopic expression of Hh. We have shown that Hedgehog can induce both stalk cell and polar cell fates in ectopic positions, probably by inducing ectopic precursor fate. Ectopic polar cells coincide with a defective oocyte anterior-posterior axis, indicative of abnormal follicle cell-oocyte signaling.

Role of Notch in stalk and polar cell lineage:
The transmembrane receptor Notch functions in a wide variety of cell fate specification events in multiple species (reviewed see ARTAVANIS-TSAKONAS et al. 1995 ). In oogenesis, Notch and its ligand Delta are involved in the fate determination of stalk and polar cells (RUOHOLA et al. 1991 ; XU et al. 1992 ; BENDER et al. 1993 ; LARKIN et al. 1996 ). Loss-of-function alleles or Notch and Delta cause a pinching-off defect in which cysts accumulate in the germarium. Analysis of enhancer trap markers for stalk and polar cell fate reveal that stalk cell formation is blocked in N^ts mutants, while polar cells accumulate in affected germaria (RUOHOLA et al. 1991 ). Expression of a constitutively active Notch construct (caN) induces extremely long, stalk-like structures, which, in the most extreme phenotype, contain neither stalk nor polar cells (LARKIN et al. 1996 ). Instead, they display markers and morphology indicative of precursors in region 3 of the germarium. Overexpression of Delta in oogenesis produces similar phenotypes (LARKIN et al. 1999 ). We have now shown that stalk and polar cells share a common lineage. Therefore, we propose that follicle cell stem cells produce precursors that are clonally distinct from other precursor cells, and divide to produce cells that assume either stalk or polar cell fate (Figure 8). The lineage relationship between stalk and polar cells clarifies previous genetic models.

Cessation of stalk and polar cell division:
Polar cells and stalk cells both stop dividing very early in oogenesis. This has been confirmed by cell morphology, clonal analysis, and visualization of mitosis (MARGOLIS and SPRADLING 1995 ; see RESULTS). Mitotic figures are not seen in mature stalk cells with DAPI staining. In addition, using PI chromosome stain in conjunction with antibody against Bib protein, we see mitotic figures in Bib-positive cells in region 3 of the germarium. Therefore, the cells are still competent to divide just before intercalation. Our data offer an alternative interpretation of data reported by MARGOLIS and SPRADLING 1995 . Based on the small size of polar cell clonal patches in stage 10 chambers, they concluded that polar cells cease division in stage 2b of the germarium, before the beginning of egg chamber pinching. Our analysis of earlier stages suggests that the patches Margolis and Spradling observed may have represented parts of bigger clones that included stalk cells, which are often difficult to detect around stage 10 egg chambers. In our model, both stalk and polar cells cease division in region 3 of the germarium.

Hedgehog can induce stalk cell and polar cell fates by affecting the precursor stage, while stall acts late in stalk formation:
The results of experiments performed with ectopic hh expression show a proliferation of follicle cells that fail to express stalk cell markers (FORBES et al. 1996A ). Our data suggest that overexpression of Hh in germarium keeps the stalk cells and polar cells in a precursor stage. Interestingly, Hh appears to induce precursor fate, even in ectopic locations, because patchy expression of Hh among the follicle cells surrounding late-stage egg chambers correlates with both ectopic stalk cell and polar cell fates. While Hh functions early in the differentiation of these cells, we find that the stall mutant (SCHUPBACH and WIESCHAUS 1991 ), which also affects stalk formation, functions downstream of the initial cell fate decision (Figure 8).

Ectopic polar cells among the follicle cells surrounding the oocyte are correlated with a defect in the oocyte anterior-posterior axis, indicative of defective follicle cell-oocyte signaling. Similar phenotypes have been observed in ovaries with patched mutant follicle cell clones (Y. ZHANG and D. KALDERON, personal communication). Therefore, as expected, ectopic Hh expression mimics patched loss-of-function phenotypes in egg chamber follicle cells.

These experiments strongly suggest that two specialized subgroups of follicle cells, stalk cells and polar cells, are related by lineage. The mutant data demonstrate that cell-cell communication affects this lineage decision. Further experiments are required to determine whether intrinsic information is delivered to one of these groups by asymmetric division. It will be interesting to analyze whether components like Numb, which are required for asymmetric cell division in neuroblasts, are required in the stalk cell/polar cell lineage.

	ACKNOWLEDGMENTS

We thank Drs. A. Spradling, T. Schüpbach, and members of the Ruohola-Baker lab for helpful comments on the manuscript. We also thank the Spradling, Schüpbach, and Struhl laboratories for providing fly lines that made this study possible. Many thanks to Ingo Ruczinski, Dr. Charles Kooperberg, and Shelley Slate for their invaluable help with the statistical analysis. This work was supported by the following grants to H.R.-B.: Established Investigatorship Award from the American Heart Association (AHA 95002990), a research grant from the March of Dimes Birth Defect Foundation (FY97-0425), National Institutes of Health (R01 HD 32464), and a Pew Memorial Trust, Pew Scholarship in the Biomedical Sciences. M.K.L. was supported by a Fellowship Cytotherapeutics Corporation.

Manuscript received July 8, 1998; Accepted for publication November 4, 1998.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ANDERSON, K., 1995  One signal, two body axes. Science 269:489-490[Free Full Text].

ARTAVANIS-TSAKONAS, S., K. MATSUNO, and M. E. FORTINI, 1995  Notch signaling. Science 268:225-232[Abstract/Free Full Text].

BASLER, K. and G. STRUHL, 1994  Compartment boundaries and the control of Drosophila limb pattern by hedgehog protein. Nature 368:188[Medline].

BECKER, H. J., 1957  Uber Rontgenmosaicflecken und Defecmutationen am Auge von Drosophila und die Entwicklungsphysiologie des Auges. Z. Induk. Abst. Vererb. Lehre 88:333-373.

BENDER, L. B., P. J. KOOH, and M. A. MUSKAVITCH, 1993  Complex function and expression of Delta during Drosophila oogenesis. Genetics 133:967-978[Abstract].

BOBOLA, N., R. P. JANSEN, T. H. SHIN, and K. NASMYTH, 1996  Asymmetric accumulation of Ash1p in postanaphase nuclei depends on a myosin and restricts yeast mating-type switching to mother cells. Cell 84:699-709[Medline].

BROWER, D. L., R. J. SMITH, and M. WILCOX, 1981  Differentiation within the gonads of Drosophila revealed by immunofluorescence. J. Embryol. Exp. Morphol. 63:233-242[Medline].

CLARK, I., E. GINIGER, H. RUOHOLA-BAKER, L. Y. JAN, and Y. N. JAN, 1994  Transient posterior localization of a kinesin fusion protein reflects anteroposterior polarity of the Drosophila oocyte. Curr. Biol. 4:289-300[Medline].

CUMMINGS, C. and C. CRONMILLER, 1994  The daughterless gene functions together with Notch and Delta in the control of ovarian follicle development in Drosophila.. Development 120:381-394[Abstract].

DOHERTY, D., L. Y. JAN, and Y. N. JAN, 1997  The Drosophila neurogenic gene big brain, which encodes a membrane-associated protein, acts cell autonomously and can act synergistically with Notch and Delta. Development 124:3881-3893[Abstract].

FORBES, A. J., H. LIN, P. W. INGHAM, and A. C. SPRADLING, 1996a  hedgehog is required for the proliferation and specification of ovarian somatic cells prior to egg chamber formation in Drosophila.. Development 122:1125-1135[Abstract].

FORBES, A. J., A. C. SPRADLING, P. W. INGHAM, and H. LIN, 1996b  The role of segment polarity genes during early oogenesis in Drosophila.. Development 122:3283-3294[Abstract].

GHO, M. and F. SCHWEISGUTH, 1998  Frizzled signalling controls orientation of asymmetric sense organ precursor cell division in Drosophila. Nature 393:178-181[Medline].

GOLIC, K., 1991  Site-specific recombination between homologous chromosomes in Drosophila.. Science 252:958-961[Abstract/Free Full Text].

GRAMMONT, M., B. DASTUGUE, and J. COUDERC, 1997  The Drosophila toucan (toc) gene is required in germline cells for the somatic cell patterning during oogenesis. Development 124:4917-4926[Abstract].

HARRISON, D. and N. PERRIMON, 1993  A simple and efficient generation of marked clones in Drosophila.. Curr. Biol. 3:424-433[Medline].

INGHAM, P. W., 1995  Signalling by hedgehog family proteins in Drosophila and vertebrate development. Curr. Opin. Genet. Dev. 5:528-534.

JAN, Y. N. and L. Y. JAN, 1998  Asymmetric cell division. Nature 392:775-778[Medline].

JOHNSON, R. and C. TABIN, 1995  The long and short of Hedgehog signaling. Cell 81:313-316[Medline].

JOHNSON, E., S. WAYNE, and R. NAGOSHI, 1995  fs(1)Yb is required for ovary follicle cell differentiation in Drosophila melanogaster and has genetic interactions with the Notch group of neurogenic genes. Genetics 140:207-217[Abstract].

KING, R. C., 1970 Ovarian Development in Drosophila melanogaster. Academic Press, New York.

LARKIN, M. K., K. HOLDER, C. YOST, E. GINIGER, and H. RUOHOLA-BAKER, 1996  Expression of constitutively active Notch arrests follicle cells at a precursor stage during Drosophila oogenesis and disrupts the anterior-posterior axis of the oocyte. Development 122:3639-3650[Abstract].

LARKIN, M. K., W.-M. DENG, K. HOLDER, M. TWOROGER, and H. RUOHOLA-BAKER, 1999  Role of Notch pathway in terminal follicle cell differentiation during Drosophila oogenesis. Dev. Genes Evol. in press.

MARGOLIS, J. and A. SPRADLING, 1995  Identification and behavior of epithelial stem cells in the Drosophila ovary. Development 121:3797-3807[Abstract].

PERRIMON, N., 1995  Hedgehog and beyond. Cell 80:517-520[Medline].

RAY, R. and T. SCHÜPBACH, 1996  Intercellular signaling and the polarization body axes during Drosophila oogenesis. Genes Dev. 10:1711-1723[Free Full Text].

RONGO, C. and R. LEHMANN, 1996  Regulated synthesis, transport and assembly of the Drosophila germ plasm. Trends Genet. 12:102-109[Medline].

RUOHOLA, H., K. A. BREMER, D. BAKER, J. R. SWEDLOWE, and L. JAN et al., 1991  Role of neurogenic genes in establishment of follicle cell fate and oocyte polarity during oogenesis in Drosophila. Cell 66:433-449[Medline].

RUOHOLA-BAKER, H., L. Y. JAN, and Y. N. JAN, 1994  The role of gene cassettes in axis formation during Drosophila oogenesis. Trends Genet. 10:89-94[Medline].

SCHÜPBACH, T. and E. WIESCHAUS, 1991  Female sterile mutations on the second chromosome of Drosophila melanogaster. II. Mutations blocking oogenesis or altering egg morphology. Genetics 129:1119-1136[Abstract].

SIL, A. and I. HERSKOWITZ, 1996  Identification of an asymmetrically localized determinant, Ash1p, required for lineage specific transcription of the yeast HO gene. Cell 84:711-722[Medline].

SPRADLING, A., 1993 Developmental genetics of oogenesis, pp. 1–70 in The Development of D. melanogaster, edited by M. BATE and A. MARTÍNEZ ARIAS. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

TAYLOR, J. R., 1997 An Introduction to Error Analysis. University Science Books, Sausalito, CA.

WOLFF, T., and D. F. READY, 1993 Pattern formation in the Drosophila retina, pp. 1277–1326 in The Development of D. melanogaster, edited by M. BATE and A. MARTÍNEZ ARIAS. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

XU, T., L. CARON, R. FEHON, and S. ARTAVANIS-TSAKONAS, 1992  The involvement of the Notch locus in Drosophila oogenesis. Development 115:913-922[Abstract].

ZAFFRAN, S., M. ASTIER, D. GRATECOS, and M. SÉMÉRIVA, 1997  The held out wings (how) Drosophila gene encodes a putative RNA-binding protein involved in the control of muscular and cardiac activity. Development 124:2087-2098[Abstract].

This article has been cited by other articles:

				E. Assa-Kunik, I. L. Torres, E. D. Schejter, D. S. Johnston, and B.-Z. Shilo Drosophila follicle cells are patterned by multiple levels of Notch signaling and antagonism between the Notch and JAK/STAT pathways Development, March 15, 2007; 134(6): 1161 - 1169. [Abstract] [Full Text] [PDF]

				M. Muzzopappa and P. Wappner Multiple roles of the F-box protein Slimb in Drosophila egg chamber development Development, June 1, 2005; 132(11): 2561 - 2571. [Abstract] [Full Text] [PDF]

				J. C. Adam and D. J. Montell A role for extra macrochaetae downstream of Notch in follicle cell differentiation Development, December 1, 2004; 131(23): 5971 - 5980. [Abstract] [Full Text] [PDF]

				S. S. Willard, E. F. Ozdowski, N. A. Jones, and C. Cronmiller stall-Mediated Extrinsic Control of Ovarian Follicle Formation in Drosophila Genetics, September 1, 2004; 168(1): 191 - 198. [Abstract] [Full Text] [PDF]

				H. R. Shcherbata, C. Althauser, S. D. Findley, and H. Ruohola-Baker The mitotic-to-endocycle switch in Drosophila follicle cells is executed by Notch-dependent regulation of G1/S, G2/M and M/G1 cell-cycle transitions Development, July 1, 2004; 131(13): 3169 - 3181. [Abstract] [Full Text] [PDF]

				K. Narbonne, F. Besse, J. Brissard-Zahraoui, A.-M. Pret, and D. Busson polyhomeotic is required for somatic cell proliferation and differentiation during ovarian follicle formation in Drosophila Development, March 15, 2004; 131(6): 1389 - 1400. [Abstract] [Full Text] [PDF]

				M. Grammont and K. D. Irvine Organizer activity of the polar cells during Drosophila oogenesis Development, March 13, 2003; 129(22): 5131 - 5140. [Abstract] [Full Text] [PDF]

				F. Besse and A.-M. Pret Apoptosis-mediated cell death within the ovarian polar cell lineage of Drosophila melanogaster Development, March 1, 2003; 130(5): 1017 - 1027. [Abstract] [Full Text] [PDF]

				F. Besse, D. Busson, and A.-M. Pret Fused-dependent Hedgehog signal transduction is required for somatic cell differentiation during Drosophila egg chamber formation Development, September 1, 2002; 129(17): 4111 - 4124. [Abstract] [Full Text] [PDF]

				J. Bai and D. Montell Eyes Absent, a key repressor of polar cell fate during Drosophila oogenesis Development, January 12, 2002; 129(23): 5377 - 5388. [Abstract] [Full Text] [PDF]

				J. R. McGregor, R. Xi, and D. A. Harrison JAK signaling is somatically required for follicle cell differentiation in Drosophila Development, January 2, 2002; 129(3): 705 - 717. [Abstract] [Full Text] [PDF]

W.-M. Deng, C. Althauser, and H. Ruohola-Baker Notch-Delta signaling induces a transition from mitotic cell cycle to endocycle in Drosophila follicle cells Development, December 1, 2001; 128(23): 4737 - 4746. [Abstract]