Developmental Biology 417 (2016) 50–62

Contents lists available at ScienceDirect

Developmental Biology
journal homepage: www.elsevier.com/locate/developmentalbiology

The Wnt pathway limits BMP signaling outside of the germline stem
cell niche in Drosophila ovaries
Violaine I. Mottier-Pavie 1, Victor Palacios, Susan Eliazer 2, Shane Scoggin 3,
Michael Buszczak n
Department of Molecular Biology, University of Texas Southwestern Medical Center at Dallas, Dallas, TX 75390, USA

art ic l e i nf o a b s t r a c t

Article history: The mechanisms that modulate and limit the signaling output of adult stem cell niches remain poorly
Received 17 March 2015 understood. To gain further insights into how these microenvironments are regulated in vivo, we per-
Received in revised form formed a candidate gene screen designed to identify factors that restrict BMP signal production to the
14 April 2016
cap cells that comprise the germline stem cell (GSC) niche of Drosophila ovaries. Through these efforts,
Accepted 26 June 2016
Available online 27 June 2016
we found that disruption of Wnt4 and components of the canonical Wnt pathway results in a complex
germ cell phenotype marked by an expansion of GSC-like cells, pre-cystoblasts and cystoblasts in young
Keywords: females. This phenotype correlates with an increase of decapentaplegic (dpp) mRNA levels within escort
Niche cells and varying levels of BMP responsiveness in the germline. Further genetic experiments show that
Stem cells
Wnt4, which exhibits graded expression in somatic cells of germaria, activates the Wnt pathway in
Wnt
posteriorly positioned escort cells. The activation of the Wnt pathway appears to be limited by the BMP
BMP
Germline pathway itself, as loss of Mad in escort cells results in the expansion of Wnt pathway activation. Wnt
pathway activity changes within germaria during the course of aging, coincident with changes in dpp
production. These data suggest that mutual antagonism between the BMP and Wnt pathways in somatic
cells helps to regulate germ cell differentiation.
& 2016 Elsevier Inc. All rights reserved.

1. Introduction terminal filament (TF) cells and cap cells, supports two to three
GSCs. Anterior escort cells (ECs) that lie immediately adjacent to
Stem cells support tissue homeostasis by providing a con- the cap cells may also assist with GSC maintenance (Rojas-Rios
tinuous source of new cells that replace cells lost to damage, dis- et al., 2012). Additional escort cells help support the early differ-
ease or normal turnover. Specialized microenvironments called entiation of germline cysts (Kirilly et al., 2011; Morris and Sprad-
niches govern the self-renewing activity of stem cells by producing ling, 2011).
a variety of factors that keep stem cells in an undifferentiated state The niche produces two bone morphogenetic protein (BMP)
(Losick et al., 2011). Stem cell daughters that remain within the family members: Decapentaplegic (Dpp) and Glass Bottom Boat
niche remain stem cells, whereas daughters displaced away from (Gbb) (Song et al., 2004; Xie and Spradling, 1998). These ligands
the niche typically undergo differentiation. The size and signaling initiate a canonical BMP signal transduction cascade by binding to
output of niches has a direct effect on stem cell number and ac- a heterodimeric receptor on the surface of GSCs, resulting in the
tivity. Therefore, changes in niche activity must be tightly regu- phosphorylation of Mothers against Dpp (Mad). pMad then binds
lated since inappropriate increases or decreases in niche function to its partner Medea and translocates into the GSC nucleus. pMad
may contribute to tumor formation and age-related declines in and Medea directly repress the transcription of differentiation
tissue homeostasis, respectively. genes, including bag of marbles (bam) and induce the expression of
The Drosophila ovary has provided a valuable platform for canonical BMP targets such as Daughters against dpp (Dad) (Chen
studying the mechanisms that influence niche formation and and McKearin, 2003a; Chen and McKearin, 2003b; Song et al.,
function in vivo. The germline stem cell (GSC) niche, comprised of 2004).
In wild-type germaria, BMP signaling rapidly decreases in GSC
daughters, known as cystoblasts (CBs), as they exit the cap cell
n
Corresponding author. niche. Recent efforts have sought to define and characterize me-
E-mail address: [email protected] (M. Buszczak).
1
Current address: KU Leuven, Belgium.
chanisms that establish this sharp gradient of BMP responsiveness
2
Current address: University of California, San Francisco. between GSCs and their differentiating daughters. Intrinsic and
3
Current address: Texas Tech University. extrinsic mechanisms, involving down-regulation of BMP pathway

http://dx.doi.org/10.1016/j.ydbio.2016.06.038
0012-1606/& 2016 Elsevier Inc. All rights reserved.
 V.I. Mottier-Pavie et al. / Developmental Biology 417 (2016) 50–62 51

components, ensure that cystoblasts displaced away from the structure subsequentially becomes branched within germline
niche experience less BMP signaling (Harris et al., 2011; Liu et al., cysts progressing through their incomplete mitotic divisions. A
2010; Schulz et al., 2002; Xia et al., 2010). However, mis-expres- substantial increase in the number of single cells with round fu-
sion of dpp, but not gbb, results in the formation of GSC-like tu- somes indicates defects in germline differentiation. Through this
mors (Song et al., 2004; Xie and Spradling, 1998), suggesting that initial small-scale screen, we found that knockdown of Wnt4 using
despite the down-regulation of pathway components (Harris et al., the c587-gal4 driver resulted in an increased number of GSC-like
2011; Xia et al., 2010), germ cells remain competent to respond to cells with round fusomes, albeit with a low penetrance (15%,
BMP ligands on some level, even as they move posteriorly through n¼120 germaria) (Fig. 1B). To confirm the RNAi phenotype, we
the germarium. next stained ovaries carrying two loss-of-function Wnt4 alleles:
Besides BMP ligands, the cap cells and escort cells of the Dro- Wnt4C1, which contains a 3 bp deletion that removes a highly
sophila germarium express a variety of other signaling molecules conserved glutamate residue at position 299 and Wnt4EMS23,
including Unpaired (Upd), Hedgehog (Hh) and Wingless (Wg) which exhibits a premature stop codon at position 343 mutants
(Decotto and Spradling, 2005; Forbes et al., 1996a, 1996b; Lopez- (Cohen et al., 2002). Both of these alleles cause partial lethality and
Onieva et al., 2008; Sahai-Hernandez and Nystul, 2013; Wang female sterility and are believed to be amorphic or strongly hy-
et al., 2008). Recent results show that the expression of Wg, a pomorphic (Cohen et al., 2002). The initial characterization of
member of the Wnt family of signaling molecules, in escort cells these alleles showed that Wnt4 mutant ovarioles were dis-
regulates the activity of follicle stem cells (Sahai-Hernandez and organized (Cohen et al., 2002). Delays in the migration of cells that
Nystul, 2013). In addition to Wg, the Drosophila genome contains a form the muscle sheet surrounding the ovarioles during devel-
number of other genes encoding Wnt ligand family members in- opment likely contribute this phenotype (Cohen et al., 2002). In
cluding Wnt2, Wnt4, Wnt6 and Wnt10, which act either through a addition to these phenotypes, we noted that germaria from
canonical pathway, involving β-catenin dependent transcriptional Wnt4C1/Df(2L)BSC291, Wnt4EMS23/Df(2L)BSC291 and from transhe-
regulation, or a non-canonical pathway (Coudreuse and Korswa- terozygous Wnt4C1/Wnt4EMS23 females also displayed an increased
gen, 2007; Logan and Nusse, 2004). Besides Wg, disruption of number of single cells with round fusomes (Fig. 1C,J). Wnt4C1/Df
Wnt4 also results in a number of morphological defects in the (2L)BSC291 and Wnt4C1/Wnt4EMS23 germaria contained large germ
ovary (Cohen et al., 2002). These phenotypes are likely caused by cell tumors (containing more than 15 single germ cells with round
defects in the apical movement of somatic cells in the developing fusomes) at 82% (n ¼103) and 85% (n ¼47) penetrance compared
gonad, marked by the disruption of the normal expression and to Wnt4EMS23/Df(2L)BSC291, which displayed large GSC/cystoblast-
distribution of FAK (Cohen et al., 2002). More recently, Wnt4 has like tumors in 25% (n¼ 77) of the germaria examined.
also been suggested to play of role in the regulation of the The Wnt4 mutants display a reduction of the number of egg
germline stem cell niche (Hamada-Kawaguchi et al., 2014; Wang chambers compared to wild type (Cohen et al., 2002). The GSC/
et al., 2015). cystoblast-like tumor phenotype, the reduction of the number of
Here we provide evidence that disruption of Wnt4 and down- egg chambers and female sterility of these alleles were rescued by
stream components of the canonical Wnt signaling pathway in a genomic HA-Wnt4 tagged construct (Fig. 1J, S1). In addition,
escort cells results in an expansion of BMP responsiveness in the expression of a Wnt4 cDNA driven by c587-gal4 also partially
germline and a subsequent increase in the number of GSCs, pre- rescued these phenotypes, supporting the model that Wnt4 is re-
cystoblasts and cystoblasts. In addition, we find loss of Wnt quired in the ECs and follicle cells to control the number of GSCs
pathway components is accompanied by an increase of dpp mRNA (Fig. 1J).
levels specifically within escort cells. Further genetic experiments Previous results showed that wingless (wg) regulates the ac-
show that Wnt4 tends to induce activation of the Wnt pathway in tivity and differentiation of follicle cells within the germarium, but
escort cells and early follicle cells of the germaria. Signaling within does not appear to greatly influence the earliest steps of germline
somatic cells of germaria appears to change during the course of development (Forbes et al., 1996b; Sahai-Hernandez and Nystul,
aging. In older flies, dpp expression within the cap cell niche de- 2013; Song and Xie, 2003). We also found that knocking down wg,
creases. This coincides with a switch in Wnt pathway activation using UAS-RNAi lines in combination with the c587-gal4 driver
from the posterior escort cells to the terminal filament and cap does not result in an appreciable germ cell phenotype, consistent
cells. These results provide new insights into how cell-cell com- with the aforementioned studies. Furthermore, RNAi knock-down
munication between specific somatic cell populations helps to of the other Drosophila Wnt ligands using the same driver did not
modulate niche signaling within the Drosophila germarium. result in germ cell phenotypes. However, we cannot exclude the
possibility that other Wnt ligands help to influence GSC differ-
entiation, as the loss-of-function achieved through RNAi knock-
2. Results down may be incomplete in our hands.
Given the phenotypes of Wnt4 mutants, we next sought to
2.1. The canonical Wnt pathway non-autonomously promotes stem determine whether loss-of-function of other canonical Wnt
cell differentiation pathway components in escort cells and follicle cells results in an
expansion of the number of single cells with round fusomes
In order to identify factors that act in escort cells to limit niche within germaria. RNAi knock-down of either frizzled (fz) and friz-
signaling and promote the differentiation of germ cells, we carried zled 2 (fz2) alone did not result in any obvious defects in the ger-
out a candidate gene RNAi screen. Targeted genes included various marium, but knock-down of both (Fig. 1D) resulted in an increased
chromatin factors and signaling molecules. We conducted the number of GSC-like cells in 65% of the germaria examined
screen by crossing available UAS-gene specific RNAi lines with the (n¼ 103). These results suggest that these receptors have re-
c587-gal4 driver, which, in adult germaria, drives expression in the dundant functions within somatic cells of the germarium, con-
escort cells and early follicle cells (Song et al., 2007). Ovaries from sistent with previous results that showed Fz and Fz2 are fully re-
the resulting females were stained for Vasa, to visualize the dundant for Wg-dependent signaling (Bhanot et al., 1999; Chen
germline, and Hts, an adducin-like protein that localizes to an and Struhl, 1999). Knock-down of arrow (also named LRP5/6 co-
endoplasmic-like organelle called the fusome. In GSCs and cysto- receptor) and dishevelled (dsh) also resulted in the formation of
blasts, the fusome (also referred to as the spectrosome in single ectopic GSC-like cells in 45% and 100% (n ¼100) of germaria re-
cells) usually appears round (de Cuevas and Spradling, 1998). This spectively (Fig. 1E,F). The knockdown of the Drosophila β-catenin
52 V.I. Mottier-Pavie et al. / Developmental Biology 417 (2016) 50–62

Fig. 1. Disruption of Wnt4 and canonical Wnt pathway components in escort cells results in the formation of GSC-like tumors. (A-I) Germaria stained for Vasa (green), Hts
(red) and DNA (blue). (A) Control germaria contain two to three GSCs. (B) Germaria from c587-gal4/þ ; UAS-Wnt4RNAi/ þ and from Wnt4EMS/Df(2 L)BSC291 displayed an
increase of GSC-like cells with round fusomes. RNAi knockdown of (D) the fz and fz2 receptors, (E) the co-receptor arrow, (F) dsh or (G) arm, using the escort cell/early follicle
cell c587-gal4 driver, results in the accumulation of GSC-like cells. (H) Over-expression of sgg in escort cells and early follicle cells also results in an expansion of un-
differentiated germ cells. (I) Overexpression of a constitutively activated form of arm (armS10) suppresses the Wnt4C1/ Wnt4EMS23 phenotype. (J) A graph illustrating the
percentage of germaria (Y-axis) that contained more (black bars) or less (light gray) than 5 round fusomes per germarium in the indicated genetic backgrounds (X axis). Note
that the GSC-like tumor phenotypes observed after disruption of the canonical pathway are rescued by a genomic Wnt4 HA-tagged construct, as well as by expression of a
Wnt4 cDNA or by expression of armS10 driven by c587-gal4. (Scale bars, 10 mm).
 V.I. Mottier-Pavie et al. / Developmental Biology 417 (2016) 50–62 53

Fig. 2. Wnt pathway mutant escort cell clones are associated with undifferentiated germ cells. (A) The MARCM system was used to generate wgRF or double fzp21 fz2C2 mutant
clones marked by GFP. Both samples were stained for GFP (green), Hts (red) and DNA (blue). Loss of Wnt pathway components in escort cells resulted in the appearance of
many single cells with round fusomes (B) Negatively marked wgRF or double fzp21 fz2C2 follicle cell clones stained for GFP (green), Hts (red) and DNA (blue). Loss of Wnt
pathway components within follicle cells did not appear to perturb early germ cell differentiation. (Scale bars, 10 mm).
54 V.I. Mottier-Pavie et al. / Developmental Biology 417 (2016) 50–62

homolog armadillo (arm) led to a similar phenotype (100%; clones did not result in an expanded number of undifferentiated
n¼ 200) (Fig. 1G). Moreover, overexpression of the kinase Shaggy germ cells (Fig. 2B). These results were consistent with our RNAi
(Sgg), which targets β-catenin for degradation, also resulted in a knock-down data and suggest that loss of Wnt signaling in escort
tumor phenotype in 94% of the germaria examined (Fig. 1H). In cells delays normal germ cell development.
addition, we observed cyst formation defects when we knock- To characterize the developmental state of the GSC-like cells in
down dsh or overexpress sgg using the c587-gal4 driver, consistent Wnt pathway mutants, we examined the expression of three
with the role of the Wnt pathway in controlling follicle cell activity molecular markers which are used to distinguish GSCs from their
(Sahai-Hernandez and Nystul, 2013). daughter cells: Nanos (Nos), Sxl (Sex-Lethal) and Bag-of-Marbles
To further test whether the canonical Wnt pathway regulates (Bam). GSCs and 16-cell cysts express Nos, a RNA binding protein
germ cell differentiation, we overexpressed a constitutively active that represses translation and helps to maintain GSCs (Forbes and
form of arm (UAS-armS10), using the c587-gal4 driver in the Wnt4 Lehmann, 1998; Kobayashi et al., 1996; Wang and Lin, 2004). All
mutant background. This mutant form of Arm escapes from ne- the GSC-like cells in Wnt4C1/Wnt4EMS23 and c587-gal4 4UAS-sgg
gative regulation by Sgg, and thus accumulates even in the ab- germline tumors express Nos at a level comparable to wild-type
sence of Wnt signal (Pai et al., 1997). When driven in the escort GSCs (n ¼80, Fig. 3A,D,G). We next analyzed the expression of Sxl
cells and early follicle cells, armS10 partially rescued the Wnt4C1/ and Bam, two differentiation-promoting factors. Wild-type GSCs,
Wnt4EMS23 tumor phenotype (Fig. 1I,J), but did not fully rescue the cystoblasts (CBs) and two cell cysts express Sxl, while cystoblasts
sterility of these transheterozygotes. In addition, expressing armS10 and early differentiating cysts express Bam. In Wnt4C1/Wnt4EMS23
rescued the tumor phenotype observed in the c587-gal4 4UAS- and c587-gal44UAS-sgg ovaries, the expression of Sxl expanded to
dshRNAi flies (Fig. 1J). These results indicate that a canonical β-ca- all the single cells with round fusomes (Fig. 3E,H). By contrast, Bam
tenin dependent pathway functions in escort cells to promote remained largely undetected in Wnt4C1/Wnt4EMS23 and c587-
germ cell differentiation downstream of the GSC. gal44 UAS-sgg germaria (only 35% and 12% of the tumorous ger-
Next, we induced homozygous mutant clones to further cor- maria displayed Bam expression in subsets of germ cells) (Fig. 3F,I).
roborate that disruption of the Wnt pathway in somatic cells of the These observations suggest that the majority of single germ cells
germarium results in expansion of undifferentiated germline cells. that arise upon Wnt pathway disruption remain in an un-
This analysis also allowed us to further map whether the Wnt differentiated GSC, pre-cystoblast-like or cystoblast-like state.
pathway acted in the escort cells or follicle cells to promote cyst
development. Because of its availability, we first generated posi- 2.2. Wnt pathway functions in adulthood to promote germ cell
tively marked mosaic analysis with a repressible cell marker differentiation
(MARCM) clones using the wgRF deficiency, which deletes a seg-
ment of genomic DNA that harbors many genes including Wnt4, To assess the extent to which the Wnt pathway functions
wg, Wnt10 and Wnt6. Escort cell clones of this deficiency resulted during development or in adulthood, we performed a series of
in the formation of germ cell tumors, marked by single cells with temporally controlled knock-down experiments. We previously
round fusomes (Fig. 2A). Next, we generated double homozygous noted that the c587-gal4 driver was less efficient at lower tem-
clones of the fzp21 and fz2C2 mutations. Consistent with the phe- peratures (Eliazer et al., 2011). We crossed the c587-gal4 driver to a
notypes seen using RNAi knock-down, loss of fz and fz2 in the line carrying UAS-armRNAi. The majority of progeny from this cross
majority of escort cells resulted in the expansion of GSC-like cells, did not exhibit a discernable phenotype when maintained at 18 °C
marked by the presence of round fusomes (Fig. 2A). Of note, wgRF during development and in adulthood (Fig. 4A,E). In contrast,
and fzp21 and fz2C2double mutant negatively marked follicle cell when female progeny resulting from this cross where kept at 29 °C

Fig. 3. Changes in germline markers upon disruption of the Wnt pathway in escort cells. Germaria of (A) control, (D) Wnt4C1/Df(2 L)BSC291 and (G) c587-gal4/þ ; UAS-sgg/ þ
stained for Hts (green) and Nos (red). (B) Control, (E) Wnt4C1/Df(2 L)BSC291 and (H) c587-gal4/ þ; UAS-sgg/ þ germaria stained for α-Spectrin (green) and Sxl (red). (C) Control,
(F) Wnt4C1/Df(2 L)BSC291 and (I) c587-gal4/þ; UAS-sgg/ þ stained for α  Spectrin (green) and Bam (red). The GSC-like cells that accumulate upon Wnt pathway disruption
express Nos and Sxl, but not the differentiation factor Bam. (Scale bars, 10 mm).
 V.I. Mottier-Pavie et al. / Developmental Biology 417 (2016) 50–62 55

Fig. 4. The canonical Wnt pathway functions during adulthood to promote germ cell differentiation. (A–D) Germaria from c587-Gal4/ þ ; UAS-arm-RNAi/ þ females raised at
(A) 18 °C, (B) 29 °C, (C) shifted from 18 °C to 29 °C upon eclosion or (D) shifted from 29 °C to 18 °C upon eclosion stained for Vasa (green), Hts (red) and DNA (blue).
Knockdown of arm in adulthood led to the formation of tumors containing more than 15 round fusomes (B and C) while knockdown of arm only during development led to
the formation of smaller tumors (D). (E) A graph showing the percent of germaria (Y axis) that contain more (black bars) or less (light gray) than 15 round fusomes per
germarium in c587-gal4/þ ; UAS-armRNAi/ þ females. (F) Graph showing the percentage of ovarioles that contain the indicated numbers of maturing egg chambers under each
of the conditions. (Scale bars, 10 mm).

both during development and in adulthood, over 80% of their the size of the germline tumors and an increase in the number of
germaria displayed a striking expansion of single undifferentiated developing egg chambers (Fig. 4D,E,F). These results indicate that
germ cells with round fusomes and a substantial decrease in the the Wnt pathway functions in adult escort cells to promote the
number of developing egg chambers (Fig. 4B,E). Next, we set c587- differentiation of germ cells. These findings also suggest that re-
gal4 4UAS-armRNAi crosses either at 18 °C, shifting the resulting ducing Wnt pathway signal transduction in c587-gal4 positive cells
progeny up to 29 °C upon eclosion, or at 29 °C, shifting the females during larval and pupal development also results in an expansion
down to 18 °C when they emerged as adults. These temperature of undifferentiated germ cells. However, loss of Wnt4, dsh or arm
shift experiments showed that knocking down arm expression does not cause escort cells to adopt a cap cell-like fate (Fig. S2).
specifically in escort cells during adulthood resulted in the for-
mation of germ cell tumors, accompanied by a reduction in the 2.3. The canonical Wnt pathway limits Dpp signaling within the
number of developing egg chambers (Fig. 4C,E,F). Similarly, re- germarium
storing arm expression after developmental knock-down resulted
in more phenotypically normal germaria, marked by a reduction in Previous studies have shown that expansion of dpp expression,
56 V.I. Mottier-Pavie et al. / Developmental Biology 417 (2016) 50–62

Fig. 5. Canonical Wnt pathway limits Dpp signaling within the germarium. (A) Control, (B) Wnt4C1/Df(2 L)BSC291, (C) c587-gal4/þ ; UAS-dshRNAi/ þ and (D) c587-gal4/ þ; UAS-
sgg/ þ ovaries carrying a Dad-LacZ enhancer trap stained for Hts (green), LacZ (red) and DNA (blue). RNA in situ hybridization of dpp and Hts immunostaning were performed
on (E) control, (F) Wnt4C1/Df(2 L)BSC291, (G) c587-gal4/þ; UAS-dshRNAi/ þ and (H) c587-gal4/þ; UAS-sgg/ þ ovaries. Arrows point to cells expressing dpp mRNA. (E) In control
germaria, dpp transcripts were mostly detected in the cap cells. (F-H) Disruption of the canonical Wnt pathway led to an upregulation of dpp transcripts in the escort cells.
RNAi knockdown of dpp partially rescued the tumorous phenotype of (I) Wnt4C1/Df(2 L)BSC291 and (J) c587-gal4/þ ; UAS-dshRNAi/ þ germaria. (K) Quantification of the
suppression of Wnt pathway mutant phenotypes by RNAi knockdown of dpp. (Scale bars, 10 mm).
 V.I. Mottier-Pavie et al. / Developmental Biology 417 (2016) 50–62 57

changes in escort cell morphology and mutations in differentiation functional significance of the observed ectopic dpp expression, we
factors can all block normal germ cell development (Kirilly et al., tested whether loss of dpp in the escort cells modified the GSC
2011; McKearin and Spradling, 1990; Xie and Spradling, 1998). expansion phenotype of the Wnt pathway mutants. Expressing
Given the loss of Bam expression and the cell autonomous func- UAS-dppRNAi in escort cells suppressed the Wnt4 mutant and c587-
tion of the Wnt pathway in escort cells, we tested whether dis- gal44 UAS-dshRNAi mutant phenotypes, so fewer germaria con-
ruption of Wnt4 and downstream pathway components resulted in tained greater than 5 single cells with round fusomes (Fig. 5I-K).
escort cell morphological defects, ectopic expression of dpp or These results together suggest that the Wnt pathway promotes
both. In germaria from young females, loss of dsh and arm or over- germ cell differentiation, at least in part, by repressing dpp ex-
expression of sgg in escort cells resulted in an expansion of single pression within the escort cells of adult ovaries.
germ cells with round fusomes despite the clear presence of some
escort cell extensions as revealed by the mCD8::GFP reporter (Fig. 2.4. Wnt signaling along the anterior-posterior axis of the
S3). However, we do observe a loss of escort cell extensions over germarium
time upon loss of Wnt pathway activity. These observations sug-
gest that the disruption of escort cell extensions by itself does not To gain a better understanding of which cells produce Wnt4
initiate the observed Wnt pathway mutant phenotypes. However and which cells respond to this ligand, we examined the expres-
changes in escort cell morphology likely do contribute to the sion of the HA-tagged Wnt4 in our rescuing genomic construct
complexity and severity of the terminal phenotype within the (Fig. S1). Staining the resulting lines for the HA epitope and Traffic
germaria of older females. Jam (Tj), a marker for somatic cells in the germarium, revealed that
Next we tested whether disruption of the Wnt pathway in es- cap cells and anterior escort cells expressed high levels of the
cort cells resulted in BMP pathway activation in germ cells outside Wnt4 tagged genomic transgene (Fig. 6A). Expression of Wnt4 was
of the cap cell niche. The Dad-LacZ enhancer trap has been ex- lower in more posteriorly positioned escort cells. The expression
tensively used as a reporter of BMP signaling in the germline (Song pattern of this tagged transgene was largely consistent with the
et al., 2004). This reporter is normally limited to the GSCs and expression of Wnt4 mRNA, as visualized by in situ hybridization
newly formed cystoblasts (Fig. 5A). In Wnt4 mutants and c587- (Fig. 6B).
gal4 4UAS-dshRNAi and UAS-sgg ovaries, the Dad-LacZ expression Next we examined the expression pattern of a previously de-
domain expands throughout the germarium (Fig. 5B-D). scribed Wnt pathway reporter (6TH) within wild-type germaria
Because the expression of Dad-LacZ may persist beyond the (Chang et al., 2008). This reporter places the β-galactosidase gene
time when a cell directly experiences BMP signaling, we also ex- under control of six TCF binding sites flanked by helper sites
amined the expression of phosphorylated Mad (pMAD), a second (Chang et al., 2008). We observed expression of this reporter in a
readout for BMP signal transduction. Like Dad-LacZ, pMAD is small number of cells within each germarium. These Wnt reporter
normally restricted to GSCs immediately adjacent to the cap cells expressing cells tended to be positioned along the escort cell and
of wild-type germaria (Fig. S4). By contrast, we observed that follicle stem cell boundary (Fig. 6C), but occasionally more ante-
within 67% of germaria (n ¼30) from c587-gal4 4UAS-dshRNAi fe- riorly positioned Wnt reporter positive escort cells were also no-
males a small subset of single germ cells with round fusomes ted (Fig. S5A). Loss of Wnt4 or dsh extinguished expression of the
outside of the normal niche expressed low levels of pMAD upon reporter in all cells while over-expression of armS10 increased the
escort cell knock-down of dsh (Fig. S4). Other single cells with number of 6TH expressing cells (Fig. S5B-D). Again these positive
round fusomes within the same tumors did not express detectable cells tended to lie along the escort cells and follicle stem cell
levels of pMAD. While these results indicate that some germ cells border, despite the expression of the armS10 transgene throughout
experience ectopic BMP signaling transduction outside the normal the escort cell population. These results suggest that unidentified
niche upon disruption of Wnt signaling, and are thus similar to mechanisms restrict the number of escort cells that respond to
GSCs, others may be slightly more differentiated and exist within a Wnt4, or that 6TH does not fully report all Wnt signaling within
pre-cystoblast or cystoblast-like state. To test this possibility, we the germarium. Double labeling experiments using the HA-tagged
examined the expression of a highly sensitive bam transcriptional Wnt4 genomic transgene and the Wnt reporter tended to show
reporter (bamP-GFP) (Chen and McKearin, 2003b) in both a wild- tight juxtaposition between cells expressing moderate levels of
type and Wnt pathway loss-of-function background. In control Wnt4 and those cells that expressed the reporter (Fig. 6C). Occa-
germaria, we first detect the GFP expression of the reporter in sionally we observed individual cells that expressed Wnt4 and
cystoblasts. This expression continued to increase as cells entered relatively high levels of the reporter. These data suggest that Wnt4
into cyst differentiation. By contrast, ectopic single germ cells may tend to act in a paracrine manner to induce a response in
present upon escort cell knock-down of Wnt pathway components more posteriorly positioned somatic cells. However, given that
expressed either no or low levels of the bam transcriptional re- Wnt4 and the reporter are expressed together in individual escort
porter (Fig. S4). Together, these results are largely consistent with cells, we cannot rule out the possibility that Wnt4 also acts in an
recently published findings (Wang et al., 2015) and suggest that autocrine fashion in certain circumstances.
loss of Wnt signaling in escort cells results in a varying degree of
ectopic BMP signal transduction in germ cells positioned away 2.5. The BMP pathway antagonizes the Wnt pathway within
from the cap cells. This, in turn, results in an expansion of un- germaria
differentiated cells that exist within a GSC-like, pre-cystoblast and
cystoblast-like state. Previous studies have shown that the BMP and Wnt signaling
To determine whether the expanded domain of Dad-LacZ and pathways antagonize each other in specific contexts, resulting in
pMAD expression reflected ectopic dpp expression, we employed a mutually exclusive domains of pathway activation (Jiang and
new, highly sensitive, in situ hybridization protocol that utilizes a Struhl, 1996). Given our observations that armS10 over-expression
series of end-labeled oligo probes. This method revealed an en- resulted in greater expression of the 6TH Wnt pathway reporter in
richment of dpp mRNA in the cap cells of wild-type ovaries posterior escort cells but not in anterior escort cells (Fig. S5), we
(Fig. 5E). Wnt4 mutants exhibited modest levels of ectopic dpp considered the possibility that Dpp expressed by the cap cells
mRNA expression in escort cells (Fig. 5F). Likewise, RNAi knock- activates the BMP signaling pathway in neighboring anterior es-
down of dsh and over-expression of sgg also resulted in an ex- cort cells of the germarium, which in turn represses Wnt signaling
pansion of dpp mRNA within escort cells (Fig. 5G,H). To test the within these cells. To test this possibility, we disrupted BMP
58 V.I. Mottier-Pavie et al. / Developmental Biology 417 (2016) 50–62

Fig. 6. Expression of Wnt4 and a Wnt pathway reporter in the germarium. (A) Germaria of transgenic flies harboring a genomic Wnt4 HA-tagged construct were co-stained
for HA (green) and Traffic jam (red). Wnt4-HA was detected at high levels in the terminal filament, the cap cells as well as in the anterior escort cells and at lower levels in
the more posterior escort cells. (B) The localization of Wnt4 mRNA (green) correlated with the Wnt4-HA staining. (C) Germaria expressing the HA-tagged Wnt4 genomic
transgene and a Wnt reporter 6TH-LacZ were co-stained for HA (green) and β-galactosidase (red). The 6TH-LacZ Wnt reporter labeled mostly escort cells and follicle stem
cells located posterior to the cells expressing the Wnt4-HA transgene. However, we did observe cases where cells displayed high levels of both Wnt4-HA and of the Wnt
reporter. (Scale bars, 10 mm).

signaling within escort cells by knocking down the expression of 3. Discussion

Mothers against dpp (Mad) using the c587-gal4 driver in combi-
nation with a UAS-MadRNAi transgene. Reduced levels of Mad re- The dynamic regulation of adult stem cell numbers and activity
sulted in an expansion of 6TH Wnt reporter expression, suggesting often depends on the size and signaling output of the specialized
that loss of BMP signaling results in ectopic Wnt pathway activa- niches within which they reside. A growing body of evidence
tion within anterior somatic cells of the germarium (Fig. 7). suggests that stem cell niches may exhibit a certain degree of
Decreased BMP signal production in aged flies contributes to a plasticity, expanding and contracting in response to local
homeostatic needs and systemic cues (Rojas-Rios and Gonzalez-
decline in GSC niche function (Pan et al., 2007). We speculated that
Reyes, 2013). Characterizing the diverse mechanisms that affect
changes in Wnt signaling response might contribute to age-related
niche size is critical for understanding the control of stem cell self-
changes in niche function to some degree. To test this idea, we first
renewal, proliferation and differentiation in vivo. Here we provide
compared dpp expression in young and aged germaria using in situ evidence that the Wnt signaling pathway, driven by the Wnt4 li-
hybridization (Fig. 8). Germaria from young females clearly ex- gand, and perhaps other Wnts as well (Wang et al., 2015), re-
hibited high levels of dpp expression in cap cells and far lower and presses the normally niche cell specific expression of dpp in the
sporadic expression in the escort cells. In contrast, cap cells from escort cells of the Drosophila ovary. Interestingly, wild-type cap
aged females display reduced levels of dpp expression while escort cells express high levels of both dpp mRNA and Wnt4, yet Wnt
cells exhibit increased expression (Fig. 8). These results are con- pathway activation is most readily seen in posterior escort cells
sistent with previous results (Rojas-Rios et al., 2012; Song et al., and early follicle cells. These results indicate that the Wnt pathway
2004). The changes in dpp mRNA expression within specific so- plays an unexpected role in regulating BMP signal production in
matic cells within the germarium correlate with changes in the somatic cells of germaria and that cells neighboring the stem cell
expression of the Wnt reporter. As noted above, in young flies microenvironment influence whether one another can produce
niche specific factors. Lastly, we show that the cells responding to
expression of the 6TH Wnt reporter is observed in posterior escort
Wnt signaling change with age, strongly correlated with a de-
cells and occasionally in anterior escort cells, but very rarely in cap
crease in niche function in older females.
cells or terminal filament cells (Figs. 6 and 7). Aged germaria ex-
hibited a dramatic change in reporter expression so that fewer 3.1. Wnt4 represses dpp expression in escort cells
escort cells expressed the reporter while an increased number of
cap cells and terminal filament cells exhibited very high levels of Drosophila Wnt4 is most closely related to vertebrate Wnts 9,
expression (Fig. 8). These results suggest that changes in both BMP 14 and 15 (Prud’homme et al., 2002) and has been shown to be
and Wnt signaling may contribute to a decline in niche activity involved in the development of dorsal epidermis, retina, heart and
during the course of aging. ovaries as well as in the left-right asymmetric morphogenesis of
 V.I. Mottier-Pavie et al. / Developmental Biology 417 (2016) 50–62 59

Fig. 7. Loss of Mad results in expanded Wnt signaling in the germarium. (A) c587-gal4 alone control and (B) c587-gal44UAS-MadRNAi knockdown germaria carrying the 6TH-
LacZ Wnt reporter were stained for β-galactosidase (red). Knockdown of Mad led to an increase of the number of anterior escort cells expressing the Wnt reporter. Arrows
point to cells expressing the 6TH reporter. (C) Graph showing quantification of the position of cells that express the 6TH Wnt reporter (c587-gal4 control n ¼ 148; c587-
gal44 MadRNAi n ¼ 135). (Scale bars, 10 mm).

the midgut (Cohen et al., 2002; Hamada-Kawaguchi et al., 2014; However, the complexity of the observed phenotypes may indicate
Inaki et al., 2007; Kuroda et al., 2012; Wu et al., 2013). Wnt4 was that loss of Wnt signaling disrupts multiple Dpp-dependent and
shown to bind to both the Fz1 and Fz2 receptors (Wu and Nusse, potentially Dpp-independent processes. For example, we observe
2002) and to signal through both the canonical and non-canonical an expansion of undifferentiated germ cells in germaria from
Wnt pathways, depending on cellular contexts (Cohen et al., 2002; young c587-gal44 UAS-dshRNAi females that still have escort cell
Hamada-Kawaguchi et al., 2014; Wu et al., 2013). Importantly, extensions, suggesting that bulk loss of escort cell protrusions does
Wnt4 has been previously shown to regulate morphological events not initiate the described phenotypes. However, over time the
during Drosophila gonad development (Cohen et al., 2002). Con- phenotype worsens, as escort cell extensions are lost, indicating
sistent with a recent report (Hamada-Kawaguchi et al., 2014), our that the ability of escort cells to maintain close contact with
data support a model in which canonical Wnt signaling functions multicellular cysts plays an important role in the continued de-
in the escort cells. However, unlike the aforementioned study, our velopment of germ cells (Kirilly et al., 2011). Ectopic expression of
findings suggest that Wnt signaling serves to limit dpp expression dpp may disrupt the normal morphology of escort cells, giving rise
within somatic cells. We observe that loss of several different to a complex phenotype involving both low levels of aberrant BMP
components of the Wnt pathway result in an expansion of single signal transduction in germ cells and disruption of other processes
cells with round fusomes, more in line with a second recently that depend on contact between escort cells and germ cells (Kirilly
published study (Wang et al., 2015). These undifferentiated cells et al., 2011; Wang et al., 2015). Future work will be needed to
express Dad-LacZ and do not express Bam protein, although further clarify how germ cells and escort cells communicate and
variable expression of pMAD and a bam transcriptional reporter influence the activity and morphology of one another.
suggests that the germ cell tumors that arise upon loss of escort
cell Wnt signaling contain a mixture of GSCs, pre-cystoblasts and 3.2. Somatic cells communicate with each other to limit niche-spe-
cystoblast-like cells. We provide evidence that dpp expression cific signaling
expands in escort cells in the absence of Wnt signaling and that
loss of dpp suppresses Wnt related phenotypes in the ovary. We observe that somatic cells of the germarium exhibit graded
60 V.I. Mottier-Pavie et al. / Developmental Biology 417 (2016) 50–62

Fig. 8. Expression of dpp and a Wnt reporter change during aging. RNA in situ hybridization of dpp was performed on germaria from females (A) aged for 4 days or (B) for
28 days. The level of dpp transcripts decreased in the cap cells and increased in the escort cells upon aging. Wnt reporter expression exhibits the opposite pattern with an
increase of Wnt activity in the cap cells and a reduction in the escort cells in the germaria of old females (D) compared to young females (C). The number of germaria
exhibiting Wnt activity either in the terminal filament (TF) and the cap cells or in the escort cells was scored in 4 d, 28 d and 35 d-old females (E). (Scale bars, 10 mm).

expression of Wnt4, from high levels in the terminal filament and non-cell-autonomous factors on stem cell aging is not limited to
cap cells to lower levels in the posterior escort cells. Despite high invertebrates. For example, parabiotic studies show that extrinsic
levels of Wnt4 expression in the anterior of the germarium, the factors from young mice can enhance stem cell activity in older
6TH Wnt reporter is most readily detected in posterior escort cells, mice (Conboy et al., 2005; Conboy and Rando, 2005). Moreover,
follicle stem cells and early follicle stem cell daughters. Signal the ability of mice to regenerate muscle, which depends on the
transduction reporters cannot always be expected to reflect the activity of satellite cells, decreases with age. This decrease in the
entirety of signaling in any given tissue because of the molecular capacity of satellite cells to respond to injury appears to be par-
and context dependent complexities that exist in vivo (Barolo, tially caused by a reduction in Notch gene expression and an in-
2006). However, despite this caveat, the expression of the 6TH crease in TGFb=b signaling, which in the muscle, represses cell
reporter in germaria appears at least responsive to levels of Wnt proliferation (Conboy et al., 2003). All of these data highlight the
signal transduction at some level since it goes away in Wnt important influence that extrinsic factors have over stem cell
pathway mutants and increases in posterior escort cells upon maintenance and activity during the course of aging across
armS10 over-expression. Interestingly, the reporter did not come on species.
in young anterior escort cells upon armS10 over-expression or upon Results presented here extend our understanding of how Dro-
armS10 over-expression coupled with TCF overexpression (data not sophila ovary niche function changes with age. High resolution
shown), despite the expression of the c587-gal4 driver throughout in situ +hybridization reveals that cap cells in young germaria
the escort cell and early follicle cell population. These observations specifically express high levels of dpp. As females age, dpp ex-
hint at mechanisms whereby Wnt pathway activation is restricted pression declines in cap cells, accompanied by a concomitant in-
in some way. The finding that loss of Mad results in expanded 6TH crease in Wnt signaling. Interestingly, we observe low levels of dpp
expression suggests that BMP signals emanating from the niche expression in aged escort cells, suggesting that the balanced mu-
attenuate Wnt pathway activation in surrounding escort cells, and tual antagonism between the BMP and Wnt pathways that exists
thereby also help to regulate escort cell function. in young escort cells becomes perturbed as these cells age. In re-
gards to aged cap cells, we speculate that increased Wnt signaling
3.3. Wnt signaling changes with age reinforces the observed loss of dpp expression. These findings
suggest that the age-dependent decline in ovarian niche function
Previous results indicate that niche function within the Droso- is not merely a passive process, marked by a general reduction in
phila ovary declines with age (Pan et al., 2007). The influence of basic cellular activity, but rather results from changes in signaling
 V.I. Mottier-Pavie et al. / Developmental Biology 417 (2016) 50–62 61

dynamics within niche cells and their close neighbors. heat-shocked as mentioned above. hs-FLP/tubgal4: UAS-GFP;;tub-
gal80, FRT2A/FRT2A and hs-FLP/tubgal4: UAS-GFP; tubgal80, FRT40A/
FRT40A were used as controls. The ovaries of the adults were
4. Materials and methods dissected 7 days after eclosion.

4.1. Fly stocks 4.4. Generation of follicle cell clones

Fly stocks were maintained at room temperature (RT) on Escort and follicle cell mutant clones for dsh3 were generated
standard cornmeal-agar medium unless otherwise indicated by by FLP/FRT-mediated mitotic recombination. 6 days after setting
experimental design. For RNAi, larvae were allowed to develop at up the crosses, larvae of the genotype hs-FLP, UbiGFP, FRT19A/ dsh3,
RT for 5 days before being moved to a higher temperature (29 °C) FRT19A were heat-shocked at 37 °C for 1 h twice a day for 3 con-
in order to increase the efficiency of knock-down. Adult flies were secutive days, and adults were dissected 7 days after eclosion. hs-
then kept at 29 °C for 5–7 days before dissection of the ovaries. As FLP, UbiGFP, FRT19A/FRT19A were used as controls. To generate
the expression of UAS-sgg driven by c587-gal4 during develop- follicle cell mutant clones for dsh3, 1–5 day old adult females of the
ment is lethal, flies were raised at RT until eclosion, adults were genotype hs-FLP, UbiGFP, FRT19A/ dsh3, FRT19A were heat-shocked
then shifted to 29 °C for a week before dissection of the ovaries. at 37 °C for 1 h twice a day for 3 consecutive days, and adults were
The following fly strains were used in this study: w1118 was dissected 14 days after the last day of heat-shock. Adults of the
used as a control; the Wnt Reporter 6TH-LacZ (MK B-31) was genotype hs-FLP, UbiGFP, FRT19A/FRT19A were used as controls. To
provided by K. Cadigan (University of Michigan); the y w hs-FLP122; generate follicle cell mutant clones for wgRF, adult females of the
Sp/Cyo; fzp21Dfz2C2 ri FRT2A/TM2 and y w hs-FLP122; wgRF, FRT40A/ genotype hs-FLP; UbiGFP, FRT40A/ wgRF, FRT40A were heat-shocked
Cyo; TM2/TM6B were a gift from G. Struhl (Columbia University); as mentioned above for the dsh3 clones and the adults hs-FLP;
c587-gal4 and Dad-LacZ lines were provided by A. Spradling UbiGFP, FRT40A/FRT40A were used as controls.
(Carnegie Institute for Science, Baltimore, MD); hs-FLP, Ubi-GFP,
FRT19A was a gift from Hugo Bellen (Baylor College of Medicine, 4.5. Immunostaining
Houston); Wnt4C1 (#6651), Wnt4EMS23 (#6650) alleles, UAS-
Wnt4RNAi (#29442), UAS-fzRNAi (#31311), UAS-fz2RNAi (#31390 and Adult ovaries were dissected, fixed, and stained according to
#27562), UAS-Arrow/LRP5/6RNAi (#31473), UAS-dshRNAi (#31306 and Eliazer et al. (2011), except that ovaries stained for pMAD were
#31307), UAS-armRNAi (#31304 and #31305), UAS-Pan/dTCFRNAi fixed for 30 min. The images were taken with a Zeiss LSM 510
(#26743), UAS-sgg (#5361), UAS-ArmS10 (#4782), Df(2 R)BSC291/ confocal microscope. The following primary antibodies were used:
Cyo (#23676), y w dsh3 neoFRT19A/FM7 (#6331), UAS-MadRNAi mouse anti-Hts (1B1 from Developmental Studies Hybridoma
(#31316) were obtained from the Bloomington Stock Center (stock Bank, dilution at 1:20), rat anti-VASA (Developmental Studies
# given in parentheses). The Wnt4::HisHA genomic construct was Hybridoma Bank, DSHB, 1:20), mouse anti-Lamin C (LC28.26,
made according to Carreira-Rosario A. et al., 2013. Briefly, an in DSHB, 1:20), anti-Engrailed (4D9, DSHB, 1:2), mouse anti-Sxl
vitro recombineering step was performed between the P[acman] (M114 DSHB, 1:10), rabbit anti-pMAD (Abcam, 1:200) mouse anti-
Bac CH321–78J12 containing the Wnt4 locus and a cassette con- BamC A7 (gift from D. McKearin, 1:20), rabbit anti-Nanos (gift from
taining the His-HA tag followed by a RFP þ marker and flanked by A. Nakamura, RIKEN, Kobe, Japan), rabbit anti-Spectrin (gift from R.
Wnt4 homology arms. Transgenic flies were then generated using Dubreuil, University of Illinois at Chicago, 1:2500), mouse anti-β-
the phiC31 integrase transgenesis system with 96E as the landing galactosidase (Promega, 1:1000), rabbit anti-GFP (Invitrogen,
site. The RFP þ cassette was then excised using the Cre enzyme 1:1000), guinea pig anti-Traffic Jam (gift from D. Godt, University
using the w; MKRS, hs-FLP86E/TM6B, Cre stock (Bloomington stock of Toronto, Toronto, ON, Canada, 1:5000), rat anti-HA (Roche,
#1501). 3F10). Cy3, Cy5, FITC (Jackson Laboratories) or Alexa 488 (Mole-
cular Probes) fluorescence-conjugated secondary antibodies were
4.2. Plasmids and rescue used at a 1:200 dilution.

To construct the pUASt-Wnt4-Flag plasmid, full-length Wnt4- 4.6. In situ hybridization

coding sequence was amplified by PCR from the RE26454 clone
(Berkeley Drosophila Genome Project), cloned into pENTR/D-TOPO In situ hybridization on adult ovaries was performed mostly as
(Invitrogen, Carlsbad, CA), and then recombined into pTWF (Ter- described in Raj A. and Tyagi S., Methods in Enzymology, 2010 for
ence Murphy, The Drosophila Gateway Vector Collection, Carnegie Drosophila imaginal discs. This method involves probing target
Institution of Washington, Baltimore, MD). The plasmid was in- mRNAs using a larger number (430) of shorter oligonucleotides
jected into w1118 embryos by Rainbow Transgenics (Camarillo, CA) (20 bases), each of which hybridize to a different portion of the
and flies were selected for the presence of the w þ transgene. target mRNA. Each of these oligonucleotides are labeled with a
Rescue of Wnt4 mutants was performed by driving Wnt4-Flag in single fluorophore at its 3′end. For designing the oligonucleotides,
ovaries with c587-gal4. we used the following web-based program (http://www.single-
moleculefish.com/, (Raj and Tyagi, 2010)). Briefly, ovaries were
4.3. Generation of escort cell clones using MARCM dissected in RNAse free 1XPBS and then fixed in 3.7% for-
maldehyde for 45 min. After being washed 2 times for 5 min with
Escort cell mutant clones for fzp21 fz2C2 and wgRF (a small de- 1XPBS, they were left in 70% EtOH overnight at 4 °C. The next day,
letion that removes Wg, Wnt4, Wnt6 and Wnt10) were generated ovaries were washed briefly in the Wash Buffer containing RNAse
by FLP/FRT-mediated mitotic recombination using the MARCM free 2XSSC and 10% of deionized formamide and then incubated
system to positively mark the clones. For the double fzp21 fz2C2 overnight at 37 °C in the dark with the dpp probe at 50 nM diluted
mutant clones, 6 days after setting up the crosses larvae of the in the Hybridization Buffer containing 2XSCC, 10% Dextran sulfate
genotype hs-FLP/tubgal4: UAS-GFP;;tubgal80, FRT2A/ fzp21 fz2C2, (Sigma D8906), 1μg/μl of yeast tRNA (Sigma R8759), 2 mM Va-
FRT2A were heat-shocked at 37 °C for 1 h twice a day for 3 con- nadyl ribonucleoside complex (NEB, S142), 0.02% RNAse free BSA
secutive days. For the wgRF mutant clones, larvae of the genotype (Ambion, AM2618) and 10% of deionized formamide. On the last
hs-FLP/tubgal4: UAS-GFP; tubgal80, FRT40A/ wgRF, FRT40A were day, ovaries were washed 2 times 30 min each 37 °C in the dark in
62 V.I. Mottier-Pavie et al. / Developmental Biology 417 (2016) 50–62

the Wash Buffer and then resuspended in a drop of Vectashield polarity genes during early oogenesis in Drosophila. Development 122,
containing DAPI (Vector Lab, H-1200). 3283–3294.
Hamada-Kawaguchi, N., Nore, B.F., Kuwada, Y., Smith, C.I., Yamamoto, D., 2014.
Btk29A promotes Wnt4 signaling in the niche to terminate germ cell pro-
liferation in Drosophila. Science 343, 294–297.
Harris, R.E., Pargett, M., Sutcliffe, C., Umulis, D., Ashe, H.L., 2011. Brat promotes stem
Author contributions
cell differentiation via control of a bistable switch that restricts BMP signaling.
Dev. Cell 20, 72–83.
V. M-P. and V. P. designed and conducted experiments, inter- Inaki, M., Yoshikawa, S., Thomas, J.B., Aburatani, H., Nose, A., 2007. Wnt4 is a local
repulsive cue that determines synaptic target specificity. Curr. Biol. 17,
preted the data and helped to write the paper. V. M-P. and V. P. 1574–1579.
carried out experiments and helped to edit the manuscript. S. E. Jiang, J., Struhl, G., 1996. Complementary and mutually exclusive activities of dec-
conducted several of the initial genetic experiments. S. S. devel- apentaplegic and wingless organize axial patterning during Drosophila leg
development. Cell 86, 401–409.
oped various transgenes and other reagents used throughout the Kirilly, D., Wang, S., Xie, T., 2011. Self-maintained escort cells form a germline stem
study and M. B. designed the experiments, interpreted the data cell differentiation niche. Development 138, 5087–5097.
and helped to write the manuscript. Kobayashi, S., Yamada, M., Asaoka, M., Kitamura, T., 1996. Essential role of the
posterior morphogen nanos for germline development in Drosophila. Nature
380, 708–711.
Kuroda, J., Nakamura, M., Yoshida, M., Yamamoto, H., Maeda, T., Taniguchi, K., Na-
kazawa, N., Hatori, R., Ishio, A., Ozaki, A., Shimaoka, S., Ito, T., Iida, H., Okumura,
Acknowledgments
T., Maeda, R., Matsuno, K., 2012. Canonical Wnt signaling in the visceral muscle
is required for left-right asymmetric development of the Drosophila midgut.
The authors would like to thank Kenneth Cadigan, Gary Struhl, Mech. Dev. 128, 625–639.
Liu, M., Lim, T.M., Cai, Y., 2010. The Drosophila female germline stem cell lineage
Allan Spradling, Hugo Bellen, Benjamin Ohlstein, the Bloomington acts to spatially restrict DPP function within the niche. Sci. Signal 3, ra57.
Stock Center and the Iowa Developmental Hybridoma Bank for Logan, C.Y., Nusse, R., 2004. The Wnt signaling pathway in development and dis-
providing reagents. Members of the Buszczak lab provided com- ease. Annu. Rev. Cell Dev. Biol. 20, 781–810.
Lopez-Onieva, L., Fernandez-Minan, A., Gonzalez-Reyes, A., 2008. Jak/Stat signalling
ments on the manuscript. This work was funded in part by the in niche support cells regulates dpp transcription to control germline stem cell
NIGMS (GM086647), the NIA (AG047318) and the E. E. and Greer maintenance in the Drosophila ovary. Development 135, 533–540.
Losick, V.P., Morris, L.X., Fox, D.T., Spradling, A., 2011. Drosophila stem cell niches: a
Garson Fogelson Endowment (UTSW Medical Center). decade of discovery suggests a unified view of stem cell regulation. Dev. Cell 21,
159–171.
McKearin, D.M., Spradling, A.C., 1990. bag-of-marbles: a Drosophila gene required
to initiate both male and female gametogenesis. Genes Dev. 4, 2242–2251.
Appendix A. Supporting information Morris, L.X., Spradling, A.C., 2011. Long-term live imaging provides new insight into
stem cell regulation and germline-soma coordination in the Drosophila ovary.
Supplementary data associated with this article can be found in Development 138, 2207–2215.
Pai, L.M., Orsulic, S., Bejsovec, A., Peifer, M., 1997. Negative regulation of Armadillo,
the online version at http://dx.doi.org/10.1016/j.ydbio.2016.06.038. a Wingless effector in Drosophila. Development 124, 2255–2266.
Pan, L., Chen, S., Weng, C., Call, G., Zhu, D., Tang, H., Zhang, N., Xie, T., 2007. Stem cell
aging is controlled both intrinsically and extrinsically in the Drosophila ovary.
Cell Stem Cell 1, 458–469.
References Prud’homme, B., Lartillot, N., Balavoine, G., Adoutte, A., Vervoort, M., 2002. Phylo-
genetic analysis of the Wnt gene family. Insights from lophotrochozoan
members. Curr. Biol. 12, 1395.
Barolo, S., 2006. Transgenic Wnt/TCF pathway reporters: all you need is Lef? On-
Raj, A., Tyagi, S., 2010. Detection of individual endogenous RNA transcripts in situ
cogene 25, 7505–7511.
using multiple singly labeled probes. Methods Enzym. 472, 365–386.
Bhanot, P., Fish, M., Jemison, J.A., Nusse, R., Nathans, J., Cadigan, K.M., 1999. Frizzled
Rojas-Rios, P., Gonzalez-Reyes, A., 2013. The plasticity of stem cell niches: a general
and Dfrizzled-2 function as redundant receptors for Wingless during Droso-
property behind tissue homeostasis and repair. Stem Cells.
phila embryonic development. Development 126, 4175–4186.
Rojas-Rios, P., Guerrero, I., Gonzalez-Reyes, A., 2012. Cytoneme-mediated delivery
Chang, M.V., Chang, J.L., Gangopadhyay, A., Shearer, A., Cadigan, K.M., 2008. Acti-
of hedgehog regulates the expression of bone morphogenetic proteins to
vation of wingless targets requires bipartite recognition of DNA by TCF. Curr.
maintain germline stem cells in Drosophila. PLoS Biol. 10, e1001298.
Biol. 18, 1877–1881.
Sahai-Hernandez, P., Nystul, T.G., 2013. A dynamic population of stromal cells
Chen, C.M., Struhl, G., 1999. Wingless transduction by the Frizzled and Frizzled2
proteins of Drosophila. Development 126, 5441–5452. contributes to the follicle stem cell niche in the Drosophila ovary. Development
Chen, D., McKearin, D., 2003a. Dpp signaling silences bam transcription directly to 140, 4490–4498.
establish asymmetric divisions of germline stem cells. Curr. Biol. 13, 1786–1791. Schulz, C., Wood, C.G., Jones, D.L., Tazuke, S.I., Fuller, M.T., 2002. Signaling from
Chen, D., McKearin, D.M., 2003b. A discrete transcriptional silencer in the bam gene germ cells mediated by the rhomboid homolog stet organizes encapsulation by
determines asymmetric division of the Drosophila germline stem cell. Devel- somatic support cells. Development 129, 4523–4534.
opment 130, 1159–1170. Song, X., Call, G.B., Kirilly, D., Xie, T., 2007. Notch signaling controls germline stem
Cohen, E.D., Mariol, M.C., Wallace, R.M., Weyers, J., Kamberov, Y.G., Pradel, J., cell niche formation in the Drosophila ovary. Development 134, 1071–1080.
Wilder, E.L., 2002. DWnt4 regulates cell movement and focal adhesion kinase Song, X., Wong, M.D., Kawase, E., Xi, R., Ding, B.C., McCarthy, J.J., Xie, T., 2004. Bmp
during Drosophila ovarian morphogenesis. Dev. Cell 2, 437–448. signals from niche cells directly repress transcription of a differentiation-pro-
Conboy, I.M., Conboy, M.J., Smythe, G.M., Rando, T.A., 2003. Notch-mediated re- moting gene, bag of marbles, in germline stem cells in the Drosophila ovary.
storation of regenerative potential to aged muscle. Science 302, 1575–1577. Development 131, 1353–1364.
Conboy, I.M., Conboy, M.J., Wagers, A.J., Girma, E.R., Weissman, I.L., Rando, T.A., Song, X., Xie, T., 2003. Wingless signaling regulates the maintenance of ovarian
2005. Rejuvenation of aged progenitor cells by exposure to a young systemic somatic stem cells in Drosophila. Development 130, 3259–3268.
environment. Nature 433, 760–764. Wang, L., Li, Z., Cai, Y., 2008. The JAK/STAT pathway positively regulates DPP sig-
Conboy, I.M., Rando, T.A., 2005. Aging, stem cells and tissue regeneration: lessons naling in the Drosophila germline stem cell niche. J. Cell Biol. 180, 721–728.
from muscle. Cell Cycle 4, 407–410. Wang, S., Gao, Y., Song, X., Ma, X., Zhu, X., Mao, Y., Yang, Z., Ni, J., Li, H., Malanowski,
Coudreuse, D., Korswagen, H.C., 2007. The making of Wnt: new insights into Wnt K.E., Anoja, P., Park, J., Haug, J., Xie, T., 2015. Wnt signaling-mediated redox
maturation, sorting and secretion. Development 134, 3–12. regulation maintains the germ line stem cell differentiation niche. Elife 4,
de Cuevas, M., Spradling, A.C., 1998. Morphogenesis of the Drosophila fusome and e08174.
its implications for oocyte specification. Development 125, 2781–2789. Wang, Z., Lin, H., 2004. Nanos maintains germline stem cell self-renewal by pre-
Decotto, E., Spradling, A.C., 2005. The Drosophila ovarian and testis stem cell ni- venting differentiation. Science 303, 2016–2019.
ches: similar somatic stem cells and signals. Dev. Cell 9, 501–510. Wu, C.H., Nusse, R., 2002. Ligand receptor interactions in the Wnt signaling path-
Eliazer, S., Shalaby, N.A., Buszczak, M., 2011. Loss of lysine-specific demethylase way in Drosophila. J. Biol. Chem. 277, 41762–41769.
1 nonautonomously causes stem cell tumors in the Drosophila ovary. Proc. Natl. Wu, J., Roman, A.C., Carvajal-Gonzalez, J.M., Mlodzik, M., 2013. Wg and Wnt4
Acad. Sci. U.S.A. 108, 7064–7069. provide long-range directional input to planar cell polarity orientation in
Forbes, A., Lehmann, R., 1998. Nanos and Pumilio have critical roles in the devel- Drosophila. Nat. Cell Biol. 15, 1045–1055.
opment and function of Drosophila germline stem cells. Development 125, Xia, L., Jia, S., Huang, S., Wang, H., Zhu, Y., Mu, Y., Kan, L., Zheng, W., Wu, D., Li, X.,
679–690. Sun, Q., Meng, A., Chen, D., 2010. The Fused/Smurf complex controls the fate of
Forbes, A.J., Lin, H., Ingham, P.W., Spradling, A.C., 1996a. hedgehog is required for Drosophila germline stem cells by generating a gradient BMP response. Cell
the proliferation and specification of ovarian somatic cells prior to egg chamber 143, 978–990.
formation in Drosophila. Development 122, 1125–1135. Xie, T., Spradling, A.C., 1998. decapentaplegic is essential for the maintenance and
Forbes, A.J., Spradling, A.C., Ingham, P.W., Lin, H., 1996b. The role of segment division of germline stem cells in the Drosophila ovary. Cell 94, 251–260.