RNAi-Based Techniques for the Analysis of Gene Function in Drosophila Germline Stem Cells

Abstract

Elucidating the full repertoire of molecular mechanisms that promote stem cell maintenance requires sophisticated techniques for identifying and characterizing gene function in stem cells in their native environment. Ovarian germline stem cells in the fruit fly, Drosophila melanogaster, are an ideal model to study the complex molecular mechanisms driving stem cell function in vivo. A variety of new genetic tools make RNAi a useful complement to traditional genetic mutants for the investigation of the molecular mechanisms guiding ovarian germline stem cell function. Here, we provide a detailed guide for using targeted RNAi knockdown for the discovery of gene function in ovarian germline stem cells and their progeny.

1 Introduction

Unlike most specialized cells, which divide to produce identical daughter cells, stem cells are endowed with the property of long-term self-renewal [1]. During cell division, a stem cell retains its unspecialized fate while producing a non-identical daughter destined for differentiation. Stem cells in the developing embryo and in some adult tissues allow for the continuous production of cells that can be instructed to differentiate into a variety of specialized fates. This unique ability, combined with recent advances in cell reprogramming and gene editing, has heightened clinical interest in the properties of stem cells [2]. Safe and effective use of stem cells and their progeny in regenerative therapies will require a detailed knowledge of the molecular factors that regulate stem cell function.

Maintenance of stem cell self-renewal is influenced by a variety of factors, including the cell’s chromatin state, metabolic status, and responsiveness to both the local microenvironment (the stem cell niche) and long-range physiological cues [3–6]. Identification and characterization of the molecular pathways that drive stem cell self-renewal thus require a model system wherein stem cells can be observed and manipulated within their native environment. While a variety of model systems have been employed to study stem cells, many of our fundamental models of stem cell function were developed as a result of experiments performed in vivo in the fruit fly, Drosophila melanogaster [7]. Like humans, Drosophila have a variety of tissue-resident stem cell populations that are critical for the continued production of specialized progeny [8–12]. The ease with which Drosophila are reared, the broad variety of genetic tools for cell-specific gene manipulation, and the high degree of conservation between many human and Drosophila cell signaling pathways make fruit flies a powerful model organism for biomedical research [13]. Drosophila are well-suited for rapid, large-scale genetic screens [14], and the availability of transgenic libraries, stock centers, and genome and bioinformatics tools [15–18] make them an efficient and cost-effective resource to study stem cell biology within the context of the whole organism.

In particular, germline stem cells (GSCs) in the Drosophila ovary have provided fundamental insights into stem cell regulation. GSCs are a specialized population of stem cells that are maintained to replenish stocks of germ cells (oocytes), whose numbers are depleted by gamete production. GSCs reside in a somatic niche composed of cap cells and terminal filament cells, located at the anterior tip of each ovariole (the functional unit of the adult Drosophila ovary; Fig. 1a) in a specialized structure called the germarium (Fig. 1b) [12, 19]. GSCs divide asymmetrically to self-renew and produce a differentiating daughter cell (cystoblast), which undergoes four subsequent rounds of mitosis with incomplete cytokinesis (Fig. 1b, c). During this process, germ cells are specified into either oocyte or nurse cell fates, and oocytes will transition into meiosis. While all germ cells in the germarium contain endoplasmic reticulum-like organelles called fusomes (Fig. 1b, c), GSCs are readily identifiable by their anteriorly localized fusome juxtaposed to the somatic cap cells [20–23]. Additional somatic cells guide the development of the germ cells. Triangular escort cells extend long projections to move the dividing cysts laterally across the germarium [24]. Follicle cells (the daughters of a second population of stem cells, follicle stem cells) envelop each germline cyst, forming an epithelial monolayer that accompanies the developing germline cyst outside of the germarium [25–27]. The resulting follicle undergoes 14 distinct phases of development, culminating in a fertilizable oocyte [19].

Fig. 1.

Drosophila ovarian germline stem cells fuel oogenesis. (a, b) The Drosophila ovary is composed of 15–20 ovarioles (one is represented in a), consisting of a germarium (b) and a progressive series of follicles in different stages of development. The germarium houses two stem cell populations: germline stem cells (GSC; dark green) and follicle stem cells (FSC; dark red). GSCs reside in a somatic niche composed of terminal filament cells and cap cells (pink). Escort cells (blue) help guide developing germline cysts posteriorly, where they are encapsulated by follicle cells (red), forming a follicle (also called an egg chamber). Germ cells (green) differentiate into oocytes (oo) or nurse cells (nc). (c) Confocal micrograph of a germarium labeled with anti-Vasa (all germ cells; green), anti-Hts (fusomes and follicle cell membranes; red), anti-LaminC (nuclear envelope of cap cells; red), and DAPI (all nuclei). Dotted lines designate GSCs. Scale bar = 10 μm

Targeted RNA interference, or RNAi, is a widely used experimental technique for gene silencing that has been successfully applied to identify novel genes that control stem cell fate and proliferation in Drosophila [28–32]. While the creation of genetic mutant alleles can be time-consuming and labor intensive, generation of RNAi against a target of interest is relatively easy and inexpensive [32–34]. Further, while complete loss-of-function genetic mutant alleles are useful tools, they may result in severe phenotypes that are difficult to interpret or developmentally lethal, precluding analysis of gene function in adult cells. In contrast, RNAi frequently reduces, but does not abolish, the targeted mRNA, resulting in hypomorphic phenotypes. By inducing RNAi using the UAS-GAL4 system, spatial control over gene knockdown in developing or adult fly tissues is readily accomplished (Fig. 2a). In this system, promoter or enhancer elements drive the expression of the yeast transcription factor Gal4 [35]. Once expressed, Gal4 binds a DNA response element, the upstream activating sequence (UAS), which precedes a gene fragment containing an inverted repeat with exact complementarity to a given gene of interest. Expression of the inverted repeat results in the formation of double-stranded RNA hairpins that trigger a sequence-specific RNAi silencing response in the cells in which it is expressed [36–39]. Thus, RNA hairpins are generated under the control of tissue- or cell-specific drivers, allowing for targeted knockdown of a gene of interest in specific cell population(s). RNAi experiments can be combined with additional genetic tools to restrict knockdown to specific temporal windows (Gal80t.s.; Fig. 2b) or to a clonal population of cells within a wild-type environment (mosaic analysis with a repressible cell marker; MARCM) [40–42]. RNAi in Drosophila ovarian GSCs thus presents an elegant experimental system to discover the molecular regulations that govern stem cell fate and function in vivo.

Fig. 2.

Targeted RNAi using the UAS-GAL4 system. (a) Virgin female tissue-specific Gal4 driver flies are crossed to male upstream activating sequence (UAS)-RNAi responder flies (UAS-IR), resulting in hairpin expression and RNAi-mediated gene knockdown in specific cells in all progeny. (b) For temporal control over Gal4 expression, Gal4 driver flies also carry a transgene in which the Gal4 inhibitor, Gal80ts, is ubiquitously expressed. In the resulting progeny, Gal4 expression is temperature-dependent. At 18 °C, Gal80^ts binds Gal4, suppressing hairpin production. At 29 °C, the conformation of the Gal80^ts protein changes, such that it can no longer bind Gal4, making Gal4 available to initiate transcription of UAS-RNAi

In this protocol, we provide an experimental strategy utilizing targeted RNAi to knock down gene function in Drosophila ovarian GSCs. This protocol can be adapted to analyze the function of any number of genes in GSCs and their progeny and is particularly useful for gene discovery. In general, the phases of the protocol include (1) selecting or creating an appropriate UAS-RNAi transgenic line against a gene of interest, (2) choosing a germline-specific Gal4 driver, (3) Drosophila husbandry to generate transgenic flies, (4) dissecting adult ovaries, (5) immunostaining, (6) image acquisition via confocal microscopy, and (7) image analysis. This basic protocol also provides a platform for further analysis of stem cell function, including gene expression, cell cycle progression, and known markers of GSC self-renewal and daughter cell differentiation. Taken together with other recently published methods [43–49], this protocol highlights the power of the Drosophila ovarian GSC as a model for the study of stem cell function in its native environment.

2 Materials

2.1 Drosophila Strains and Culture

1. Drosophila strains (see Note 1), including a germline-compatible UAS-RNAi responder strain and a germline-specific Gal4 strain.

2. Standard fly culture medium (cornmeal/molasses/yeast/agar) in bottles and vials (see Note 2).

3. Yeast paste: 1–1 mixture of dry active yeast (see Note 3) and distilled water, mixed to the consistency of smooth peanut butter. Store covered with Parafilm at 4 °C to prevent drying.

4. Dissecting stereomicroscope with dual goose-neck illumination, 0.6–4× magnification range, and a flat black dissection base, equipped with a fly pad and air needle to deliver CO₂ (Drosophila anesthetic; see Note 4).

2.2 Ovary Dissection and Immunostaining

1. 1.5 mL microfuge tubes, pre-coated (see Note 5).

2. 15-mL polypropylene centrifuge tubes.

3. 50-mL polypropylene centrifuge tube.

4. Glass or plexiglass dissection dish.

5. Kimwipes.

6. Glass Pasteur pipettes and bulbs.

7. Two pairs of sharpened forceps (INOX, Dumont #5, Biologie point).

8. Two 27 × 1¼ gauge needles with 1 mL syringes.

9. Cold room-safe orbital nutator (also called a GyroMixer).

10. Grace’s Insect Medium without additives (Lonza; see Note 6).

11. Phosphate-buffered saline (PBS).

12. Wash solution: 0.1% Triton-X-100 in PBS.

13. Fixative: 5.3% formaldehyde in Grace’s media (see Note 7). Prepare fresh prior to dissection. For each sample, add 300 μL 16% formaldehyde (Ted Pella; see Note 8) and 600 μL Grace’s media. Keep on ice.

14. Blocking Solution: 5% bovine serum albumin (BSA), 5% normal goat serum, 0.1% Triton-X-100 in PBS (see Note 9). Prepare using sterile technique, and store at 4 °C. Discard if cloudy.

15. Primary Antibodies: In nearly all initial experiments, we use two primary antibodies that allow for identification of germ-line and somatic cells in the ovary (see Fig. 3c, d). Mouse anti-Hts (1B1-s; Developmental Studies Hybridoma Bank) is diluted in blocking solution to a final concentration of 1:10 and labels both the germline-specific fusome and follicle cell plasma membranes. Mouse anti-LaminC (LC28.26-s; Developmental Studies Hybridoma Bank) is diluted in blocking solution to a final concentration of 1:100 and strongly labels the nuclear envelope of cap cells (see Notes 10 and 11).

16. Secondary Antibodies: Species-matched secondary antibodies tagged with a fluorophore of interest. For the detection of anti-Hts and anti-LaminC (see Fig. 3c, d), we use Alexa Fluor goat anti-mouse-568 (Life Technologies; see Note 12), diluted to a final concentration of 1:200 in blocking solution.

17. 4′,6-Diamidino-2-phenylindole (DAPI): For working concentration, prepare a 1:500 dilution of a 5.0 mg/mL DAPI stock solution (see Note 13) in 0.1% Triton-X-100 in PBS.

18. Mounting Media: 20 mg/mL n-propyl gallate in glycerol (see Note 14). In a 50 mL conical tube, combine 1.0 g n-propyl gallate (Sigma) with 5 mL PBS; vortex to mix. Add 45 mL 100% glycerol. Cover the conical tube in foil and rotate on a nutator at room temperature overnight. Media should be clear but very viscous. Store protected from light at 4 °C. A working stock can be poured into an opaque dropper bottle for ease of application.

Fig. 3.

Germline-targeted RNAi knockdown of bam results in small ovaries filled with undifferentiated germ cells. (a, b) Bright-field images of intact control (a) and germline-specific bam knockdown (b) ovaries 5 days after eclosion. (c, d) Confocal micrographs of representative ovarioles labeled with anti-Hts (fusomes and follicle cell membranes; red), anti-LaminC (nuclear envelope of cap cells; red), and DAPI (all nuclei). While nos-Gal4 control germaria (c) contain GSCs (dashed line), cystoblasts (filled arrowheads), and differentiating germ cells (recognized by branched fusomes; open arrowheads), bam knockdown germaria have accumulated GSC- and cystoblast-like cells throughout the germarium, at the expense of differentiated germ cells. Further, bam knockdown results in a block to oogenesis, as evidenced by the lack of progressively developed follicles (d’) typically found in control ovarioles (c’). Scale bar = 200 μm (a, b), 10 μm (c, d)

2.3 Sample Preparation and Confocal Microscopy

1. Two needle holders with sharpened tungsten needles; one should have an “L”-shaped bend at the tip (see Note 15).

2. Glass microscope slides and coverslips (1 μm thickness, 22 × 22 mm).

3. Steel weight, measuring approximately 250 g.

4. Fingernail polish (Sally Hansen Tough as Nails, any color).

5. Laser scanning confocal microscope, inverted, equipped with 63× oil immersion lens (n.a. = 1.4) and 1.5–3× optical zoom (see Note 16).

6. Confocal image acquisition and analysis software (see Note 17).

3 Methods

3.1 Drosophila Strains and Culture

3.1.1 Selecting a Gene-Specific UAS-RNAi Responder Transgene

Although a wide variety of target-specific UAS-RNAi transgenes are available in Drosophila [32], care must be taken to ensure successful knockdown of your gene of interest in female GSCs. A variety of commonly used Drosophila genetic tools, including the UASt responder lines that promote strong expression in somatic cells, do not work effectively in the female germline [39, 40, 50, 51]. To circumvent this problem, Ni and colleagues developed two RNAi vectors (Valium20 and Valium22) specifically designed for optimal expression of RNAi hairpins in the germline [50]. Valium20 promotes strong knockdown in both the ovarian germ-line and soma; in contrast, the Valium22 vector, based on the germline-compatible UASp vector [51], promotes strong expression in the germline, but only moderate to weak expression in the soma [50, 52]. A large collection of target-specific Valium20- and Valium22-based UAS-RNAi transgenic lines (henceforth referred to as “UAS-RNAi responders”) are now available from the Transgenic RNAi Project at Harvard Medical School (TRiP; www.flyrnai.org; see Note 18) via the Bloomington Drosophila Stock Center (see Note 1; Table 1). Another large collection of UAS-RNAi responders are available from the Vienna Drosophila Stock Center (VDRC). The VDRC maintains two libraries of UAS-RNAi responders that differ primarily in the insertion site of the UAS-RNAi into the Drosophila genome [36]. Although neither of these libraries were created in germline-optimized vectors, reliable expression of gene-specific RNAi hairpins has been achieved in the germline via co-expression of Dicer2 (UAS-Dcr2) [29, 30].

Table 1.

Resources for RNAi-mediated gene knockdown in Drosophila ovarian GSCs and somatic support cells

	Transgenes	Source	Adult ovarian expression pattern	References
Gal4 drivers
Germline	nos-Gal4::VP16 (III)a	BLMb #4937	Moderate expression in GSCs, strong expression in germline cysts	[51, 55]
	nos-NGT40; UAS-Dcr2	BLM #25751	Very weak expression in GSCs	[29, 30]
	MTD-Gal4	BLM #31777	Moderate expression in GSCs, very strong expression in germline cysts	[54]
	ET-Flpx2Gal4–168	Drosophila community	Germline clones	[41]
Niche	bab1-Gal4	BLM #6802; Drosophila community	Cap cells and escort cells	[61, 62]
	c587-Gal4	Drosophila community	Escort cells, follicle stem cells, prefollicle cells	[69]
	Tj-Gal4	KYOc #104055	Terminal filament & cap cells	[70]
	ET-Flpx2Gal4-688A	Drosophila community	Escort cell clones	[41]
	13C06-Gal4	BLM #47860	Escort cells, follicle stem cells, and prefollicle cells	[71]
	ET-Flpx2Gal4-398A	Drosophila community	Escort cells, follicle stem cells, and prefollicle cell clones	[41]
UAS-RNAi responders
Long hairpins	VDRC KK	VDRCd	Germline expression with UAS-Dcr2	[36]
	VDRC GD	VDRC	Germline expression with UAS-Dcr2	[36]
Short hairpins	TRiPe Valium10	BLM	Expression limited to somatic cells	[39]
	TRiP Valium20	BLM	Expression in somatic and germline cells	[50]
	TRiP Valium22	BLM	Strong expression in germline cells, weaker in somatic cells	[50]

^aNote that a second insertion is available on the X chromosome; however, its expression is weaker than the insertion on the III

^bBloomington Drosophila Stock Center at Indiana University, Bloomington, IN, USA

^cKyoto Stock Center at the Kyoto Institute of Technology, Kyoto, Japan

^dVienna Drosophila Resource Center, Vienna, Austria

^eTransgenic RNAi Resource Project at Harvard Medical School, Boston, MA, USA

While an ever-increasing number of germline-compatible UAS-RNAi responders are being made public (see FlyBase, www.flybase.org, for a current listing), these tools may not yet be available for some genes of interest. In this case, it is possible to create a Valium20- or Valium22-based UAS-RNAi transgene using standard molecular cloning techniques and have this transgene commercially injected into embryos for the establishment of transgenic lines (see Note 19). While creation of customized UAS-RNAi lines does require time to establish new transgenic lines, it is a viable and cost-effective option for genes with few existing tools. Detailed protocols, cloning vectors, and primer sequences are available from the TRiP (www.flyrnai.org) [50].

Three common issues arise when using targeted UAS-RNAi approaches. First, the strength of knockdown of a gene of interest can vary between UAS-RNAi transgenes. This can be assessed by measuring RNA or protein levels using a variety of techniques but is perhaps best visualized in vivo by immunofluorescence against the protein of interest if suitable reagents are available. Second, a given RNAi hairpin may target multiple genes, resulting in phenotypes not due to loss of the gene of interest (“off-target effects”). This is particularly true for the VDRC KK library [53]. To minimize the chances of off-target effects, choose more than one hairpin, targeted at different regions of the gene of interest, for RNAi analysis. The most effective control is to design hairpins with high sequence similarity to the experimental hairpin, but with several mismatched bases, thus precluding the induction of RNAi [33, 34]. Lastly, for genes that encode multiple isoforms, hairpins could target one, many, or all isoforms of the mRNA, depending on the targeted region. It is important to recognize the location of the hairpin sequence with regard to the predicted transcript, as this could impact the interpretation of results.

3.1.2 Selecting a Germline-Specific Gal4 Driver Transgene

As in the case of the UAS-RNAi responder, a germline-compatible Gal4 driver must also be selected to achieve gene-specific knockdown in GSCs. As mentioned above, ubiquitous expression of a UAS-RNAi responder could cause developmental lethality, thus precluding analysis of adult GSC phenotypes. Further, some drivers that are considered to be “ubiquitous” (such as the hsGal4 used in many somatic experiments) do not work well in the germline [35, 40, 51]. Thus, a germline-specific Gal4 driver is required to restrict UAS-RNAi responder expression to GSCs. Unfortunately, very few germline-specific drivers exist, and even fewer drive expression robustly in GSCs; however, a few lines can be used to successfully knock down gene expression in GSCs and their daughters without directly modifying the surrounding somatic cells (Table 1).

Perhaps the most frequently used germline drivers in the Drosophila ovary are the nos-Gal4::VP16 and the “maternal triple” MTD-Gal4 [43, 51, 54, 55]. Two important caveats should be noted, however, when employing the nos-Gal4::VP16 driver in GSC experiments. First, nos-Gal4::VP16 expression is not limited to adult GSCs [51, 55]. This driver is expressed in the developing primordial germ cells, larval/pupal germ cell precursors, adult GSCs, and nearly all GSC progeny. It is therefore difficult to conclusively demonstrate that a given gene of interest functions solely in adult GSCs (i.e., GSC function versus GSC establishment). (Unfortunately, the Gal80^t.s. system cannot help in this regard, as Gal4::VP16 lacks the Gal80-binding site [56].) Second, while nos-Gal4::VP16 is expressed in adult GSCs, it is expressed at much higher levels in differentiating daughter cells (particularly 2-, 4-, 8-, and 16-cell cysts) [51, 55]. Thus, any knockdown experiments that result in phenotypes impacting germline differentiation must be carefully interpreted. For example, nos-Gal4::VP16-mediated knockdown of the differentiation factor bag of marbles (bam) using targeted RNAi results in a block in germline differentiation, leading to the accumulation of single germ cells and 2-cell germ cysts at the expense of 4-, 8-, and 16-cell cysts (Fig. 3). As an isolated experiment, these results could indicate that bam is required for the differentiation of 2-cell cysts; however, analysis of Bam protein expression and a bam genetic null allele clearly demonstrates that bam is required for the initial differentiation of the immediate GSC daughter cell (the cystoblast) [57–60]. Similar considerations should be made when using MTD-Gal4, a compound driver consisting of three independent Gal4 insertions (nos-Gal4::VP16 plus pCOG-Gal4::VP16 and nos.NGT40-Gal4) [43, 51, 54]. Intriguingly, neither pCOG-Gal4::VP16 (also called otu-Gal4::VP16) nor nos.NGT40-Gal4 are highly expressed in adult GSCs (E. Ables unpublished and Ref. [51]); however, several research groups have obtained GSC-related phenotypes when using this driver (in combination with UAS-Dcr2) in large-scale screens [29, 30], indicating that low levels of transgene expression are sufficient to induce phenotypes in some situations. Finally, a newly characterized Gal4 driver, ET-Flpx2Gal4-398A, is also available for clonal expression of UAS-RNAi responders in the germ-line [41], but has not yet been widely utilized.

A variety of somatic Gal4 drivers are also available that can be used to identify microenvironmental factors that influence GSC function (Table 1). Most of these somatic Gal4 drivers robustly promote the expression of a wide variety of UAS-RNAi transgenes (except the Valium22 series). Researchers should note, however, that many of these drivers are expressed in overlapping populations of somatic cells and may also be expressed outside of the ovary. For example, bab1-Gal4 is frequently used to drive UAS-RNAi expression in the somatic cap cells of the ovary, but it also shows strong expression in the somatic polar and stalk cells present in more developed stages of oogenesis [61, 62]. Researchers should test driver expression in multiple tissues to preclude non-autonomous effects prior to any knockdown experiment and check the expression pattern of the driver over the course of the experiment to ensure the pattern does not change due to the genetic manipulation or nutrient status of the fly.

3.1.3 Drosophila Husbandry

1. Maintain fly cultures using standard procedure (see Note 20).

2. Culture at least two bottles each of the Gal4 driver line and an isotype control (typically yw; see Note 21) and one bottle of the UAS-RNAi responder line (see Note 22). Collect virgin females from the Gal4 driver line and the isotype control.

3. Set control (Gal4 driver with yw males; UAS-RNAi responder with yw female virgins) and experimental (Gal4 driver with UAS-RNAi responder) crosses (Fig. 2a) in vials at a density of five pairs of flies per vial (see Note 23). Crosses should be set in replicates. Addition of dry yeast to the culture medium provides optimal conditions to increase egg laying. Maintain vials at 25 °C, transferring adults to new vials every 3 days. New flies should emerge within 10 days of this date (at this temperature). For experiments using a Gal80ts repressor (Fig. 2b), crosses should be maintained at 18 °C, and new flies will emerge within 18–20 days.

4. Allows crosses to eclose for 1–2 days. Collect the female flies you will dissect and an equal number of accompanying males (typically 12–15 pairs of flies per vial), and add them to fresh vials supplemented with wet yeast paste (see Note 24).

5. Transfer adults into freshly yeasted vials daily, and maintain at a constant temperature (typically 25 °C or 29 °C; see Note 25) until the day of dissection. For experiments using a Gal80ts repressor, flies should be maintained at 29 °C. The age of the flies at dissection (determined by the number of days post-eclosion) is determined by the experimenter. For the analysis of GSC self-renewal, it is appropriate to collect flies at a series of timepoints (i.e., 0 days, 4 days, 8 days, and 12 days after eclosion) to chart GSC maintenance over time. The timing of dissection post-eclosion will depend on the number of lines being screened and the exact phenotypes expected/discovered and should be determined experimentally.

3.2 Ovary Dissection and Immunostaining (Fig. 4)

Fig. 4.

Experimental work-flow for ovary immunostaining. Preparation of Drosophila ovaries for confocal imaging is outlined, highlighting the addition of primary and secondary antibodies and nuclear counterstain

1. Use CO₂ to anesthetize flies. Remove males and discard. Dissect ovaries from female flies in cold Grace’s medium in a glass or plexiglass dissecting dish. To dissect ovaries (see Note 26), grasp the anterior end of the female’s abdomen with forceps in the left hand. Using forceps in the right hand, pinch the last stripe in the female’s abdomen and pull away from the body. The forceps will grab not only the cuticle but also the urogenital system (frequently including the ovaries). Repeat dissection for all of the females in that vial. Collect dissected ovaries in cold Grace’s medium, and quickly break open the outer muscle sheath surrounding the ovarioles to separate individual ovarioles. Using a glass Pasteur pipet, remove BSA from a labeled, pre-coated microfuge tube and discard. Using the same pipet (which is now also coated with BSA), move the dissected ovaries in Grace’s medium to the microfuge tube. Place the tube on ice.

2. Repeat dissection with additional vials of flies. Proceed with fixation (below) within an hour of dissection of the first fly.

3. Allow ovaries to collect in the bottom of the microfuge tube by gravity. Remove as much Grace’s media as possible (see Note 27), and add 1.0 mL of Fixative. Invert the microfuge tube several times to ensure all ovaries are suspended, and rotate on a nutator for 13 min at room temperature.

4. Remove fixative to an appropriate waste container, and quickly add 1.0 mL of wash solution. Invert the tube several times, and remove the wash to the waste container. Add fresh wash solution, invert the tube to suspend the ovaries, and rotate on a nutator for 15 min.

5. Discard the solution, and wash ovaries two more times in 1.0 mL of wash solution for greater than 10 min.

6. Discard the solution, and add 1.0 mL of blocking solution. Rotate ovaries for 3 h on nutator at room temperature.

7. Discard the solution, and add 400 μL of primary antibody mixture (see Notes 28 and 29). Incubate ovaries overnight or over two nights at 4 °C on a nutator.

8. Remove antibody (see Note 30) and wash ovaries four times in 1.0 mL wash solution for at least 30 min each on a nutator.

9. Discard the last wash, and add 500 μL of secondary antibody mixture. Incubate ovaries for 1–2 h at room temperature on a nutator. From this step forward, all samples should be protected from light to minimize loss of fluorescence.

10. Discard antibody and wash ovaries four times in 1.0 mL wash solution for at least 30 min each on a nutator.

11. Discard the last wash, and add 500 μL of DAPI. Incubate ovaries for 15 min on a nutator.

12. Discard DAPI, and wash ovaries two times in 1.0 mL wash solution for 10 min each on a nutator.

13. Discard the last wash, removing as much liquid as possible. Add four drops of mounting media directly on top of the ovaries. Samples can be stored upright at 4 °C for up to a month.

3.3 Sample Preparation and Confocal Imaging

1. Prepare samples on glass slides for imaging on the day of or day before imaging (see Note 31).

2. Using a glass Pasteur pipet, move ovaries and mounting media onto a slide. Remove excess mounting media. Isolate (tease) individual ovarioles apart. For optimal imaging of GSCs, remove all large follicles (see Note 32). Spread ovarioles out across the slide, making sure there are no large clumps.

3. Once all large follicles are removed and ovarioles are sufficiently spread away from each other in the center of the slide, add 1–2 small drops of mounting media back to slide. Drop a glass coverslip onto the center of the slide, and allow capillary action to spread the mounting media and ovarioles across the coverslip. If any air remains between the coverslip and the slide, use a Pasteur pipet to slowly add a little more mounting media at the edges of the coverslip.

4. Place Kimwipes and a steel weight on top of the coverslip/slide to flatten the ovarioles. After 10–15 min, remove the weight and seal the coverslip to the slide using fingernail polish. Allow slides to dry before imaging.

5. Use an inverted confocal microscope with a 63× oil immersion lens and an optical zoom (1.5–3×) to image samples. Images are collected as confocal z-stacks (1 μm optical sections). Each germarium on the slide should be imaged for analysis (typically 75–100 germaria per slide).

3.4 Phenotypic Analysis, Quantification, and Next Steps

RNAi-mediated knockdown of genes that function in the Drosophila germline may result in a variety of phenotypes. Loss of expression of genes that function in an early step of germline establishment (i.e., during development) or germline differentiation may result in a complete block to oogenesis. Knockdown of genes that maintain the GSC fate may result in a complete loss of GSCs (and, thus, all daughter germ cells), resulting in rudimentary ovaries devoid of germ cells. Loss of expression of genes that function in early cyst growth (independent of GSC function or germ-line differentiation) may result in ovaries lacking more developed follicles. Although each of these possible scenarios results in a similar “small ovary” phenotype, knockdown of some genes may result in subtle phenotypes that do not alter overall ovarian morphology. It is therefore important to perform a thorough characterization of ovariole structure and cellular composition as a starting point to more detailed molecular analyses.

Following the immunofluorescent detection of anti-Hts and anti-LaminC, early germ cells can be recognized by the presence of the Hts-positive fusome (see Fig. 3c, c’). GSCs are unambiguously identified via their anteriorly localized fusome adjacent to the cap cells (which express high levels of LaminC). While the fusome is typically round in shape, changes in fusome morphology (elongation and fragmentation) correlate with the different phases of the GSC cell cycle and can be used as an indicator of cell cycle progression [20, 23]. When the GSC divides to form a cystoblast, a new fusome forms, located in the posterior of the cystoblast but with a similar round shape as the GSC fusome (arrowheads, Fig. 3c). As the cystoblast continues to divide with incomplete cytokinesis to form multicellular germline cysts, the fusome branches such that a single branch endpoint is located in each cell cytoplasm (open arrowheads, Fig. 3c). Initial quantification of germ cell number at each stage of differentiation (GSCs, cystoblasts, 2-cell, 4-cell, 8-cell cysts) can thus provide a snapshot of GSC maintenance and the ability of the daughter germ cells to divide and differentiate.

Alterations in GSC or cystoblast number are the initial phenotypic indicators of a change in GSC function. Reduction or complete loss of GSCs (quantified as the average number of GSCs per germarium) is indicative of a failure to maintain the GSC population. Additional analyses can then be performed to elucidate the cause of GSC loss. For example, apoptosis assays (such as immunofluorescence for TUNEL or cleaved caspase) can indicate whether GSCs are lost due to cell death, and molecular markers for each of the differentiated stages can be assessed to determine whether GSCs are lost due to premature differentiation. Such markers may include phosphorylated Mothers Against Decapentaplegic (pMad, highly expressed in GSCs), Bam (expressed in cystoblasts and 2- and 4-cell cysts), Ataxin-2-binding protein 1 (A2BP1, expressed in 4- and 8-cell cysts) [63], and C-3-G and Orb (expressed in oocytes). Expansion of the number of GSCs and/or cystoblast-like cells per germarium may indicate an increased rate of GSC proliferation or a failure of the immediate daughter cell to initiate differentiation. To evaluate differences in GSC proliferation, the S-phase indicator 5-ethynyl-2′-deoxyuridine (EdU) or the M-phase indicator phospho-histone H3 can be added to the analysis. These markers are also useful in examining changes in cystoblast differentiation and proliferation. Immunofluorescence analysis in targeted RNAi knockdown ovaries can be complemented by molecular analysis of the ovary, including gene expression (RT-PCR, RNA-seq, or in situ hybridization) or protein analysis (Western blotting). All phenotypes derived from targeted RNAi should be validated using other gene loss-of-function models, including the phenotypic assessment of genetic mutants by mosaic clonal analysis [45, 48].