Interaction between Differentiating
Cell- and Niche-Derived Signals in
Hematopoietic Progenitor Maintenance
Bama Charan Mondal,1,6Tina Mukherjee,1,6Lolitika Mandal,1,6,7Cory J. Evans,1Sergey A. Sinenko,1
Julian A. Martinez-Agosto,2,3,5,* and Utpal Banerjee1,3,4,5,*
1Department of Molecular, Cell, and Developmental Biology
2Department of Human Genetics, David Geffen School of Medicine at UCLA
3Molecular Biology Institute
4Department of Biological Chemistry
5Eli and Edythe Broad Center of Regenerative Medicine and Stem Cell Research
University of California Los Angeles; Los Angeles, CA 90095, USA
6These authors contributed equally to this work
7Present address: Indian Institute of Science Education and Research, Mohali, India
*Correspondence: firstname.lastname@example.org (J.A.M.-A.), email@example.com (U.B.)
Maintenance of a hematopoietic progenitor popu-
lation requires extensive interaction with cells
hematopoietic organ, niche-derived Hedgehog sig-
naling maintains the progenitor population. Here,
we show that the hematopoietic progenitors also
require a signal mediated by Adenosine deaminase
growth factor A (Adgf-A) arising from differentiating
cells that regulates extracellular levels of adeno-
sine. The adenosine signal opposes the effects of
Hedgehog signaling within the hematopoietic pro-
genitor cells and the magnitude of the adenosine
signal is kept in check by the level ofAdgf-A secreted
from differentiating cells. Our findings reveal signals
arising from differentiating cells that are required for
maintaining progenitor cell quiescence and that
function with the niche-derived signal in maintaining
the progenitor state. Similar homeostatic mecha-
nisms are likely to be utilized in other systems that
maintain relatively large numbers of progenitors
that are not all in direct contact with the cells of the
The mammalian hematopoietic niche displays complex interac-
tions between populations of HSCs and progenitors to maintain
their numbers (Garrett and Emerson, 2009). The relative in vivo
contributions of cues emanating from the microenvironment in
regulating stem cell versus progenitor maintenance remains
unclear (He et al., 2009; Martinez-Agosto et al., 2007; Orkin
and Zon, 2008). Several (Jude et al., 2008; Tumbar et al.,
2004), but not all (Li and Clevers, 2010) stem cell and progenitor
populations demonstrate slow cell cycling and this property
of ‘‘quiescence’’ is critical for maintaining their integrity over
a period of time.
In vivo genetic analysis in Drosophila allows for the study
of stem cell properties in their endogenous microenvironment
(Losick et al., 2011). Drosophila blood cells, or hemocytes,
develop within an organ called the lymph gland, where differen-
microenvironment or niche, are found (Jung et al., 2005). Differ-
entiated blood cells in Drosophila are all myeloid in nature and
are located along the outer edge of the lymph gland, in a region
termed the cortical zone (CZ; Jung et al., 2005; Figure 1A). These
arise from a group of progenitors located within an inner core of
cells termed the medullary zone (MZ). The MZ cells are akin to
the common myeloid progenitors (CMP) of the vertebrate hema-
topoietic system. They quiesce, lack differentiation markers, are
multipotent, and give rise to all Drosophila blood lineages (Jung
et al., 2005; Krzemien et al., 2010). MZ progenitors are main-
tained by a small group of cells, collectively termed the posterior
signaling center (PSC), that function as a hematopoietic niche
(Crozatier et al., 2004; Krzemien et al., 2007; Mandal et al.,
2007). Clonal analysis has suggested the existence of a niche-
bound population of hematopoietic stem cells (Minakhina and
Steward, 2010), although such cells have not yet been directly
The PSC cells express Hedgehog (Hh), which is required for
the maintenance of the MZ progenitors (Mandal et al., 2007).
Cubitus interruptus (Ci) is a downstream effector of Hh signaling
form in the presence of Hh and degraded to the repressor Ci75
form in the absence of Hh (Smelkinson and Kalderon, 2006).
PSC-derived Hh signaling causes MZ cells to exhibit high Ci155
(Mandal et al., 2007).
Proliferation of circulating larval hemocytes is also regulated
by Adenosine Deaminase Growth Factor-A (Adgf-A), which is
Cell 147, 1589–1600, December 23, 2011 ª2011 Elsevier Inc. 1589
Figure 1. Regulation of Progenitor Quiescence by Differentiating Hemocytes Mediated by Pvf1/Pvr and Adgf-A Signaling
(A) A representative lymph gland primary lobe from a midsecond instar larva (left image) consisting of a posterior signaling center (PSC, yellow), progenitors
(green), and a few differentiating cells (red). A primary lobe from a later-staged wandering third instar larva (right image) shows three distinct zones: the PSC
(yellow), which functions as the niche for the maintenance of progenitors (green) within the medullary zone (MZ), and the peripheral cortical zone (CZ) region,
comprised of differentiating blood cells (red).
(B–G) Cell proliferation profile in lymph glands from midsecond instar larvae analyzed by BrdU (red) incorporation. (B) Control lymph glands (genotype: Hml-gal4)
at this stage have only a few proliferating cells (9 ± 6 cells, n = 6). (C) Induction of cell death in the differentiating cells by expression of Hid and Reaper (Hml-
gal4 UAS-hid UAS-rpr) causes an increase in the number of BrdU positive cells (red), indicative of the loss of quiescence among the progenitors (compare with
[B]). A similar loss of quiescence occurs upon (D) downregulation of Pvf1 in the PSC using Pvf1RNAi(Antp-gal4 UAS-2xEGFP UAS-Pvf1RNAi), (E) downregulation
of Pvr in the differentiating cells using PvrRNAi(Hml-gal4 UAS-PvrRNAi; 39 ± 7 cells, n = 5, p < 0.004 when compared to (B), or (F) downregulation of Adgf-A in
the differentiating cells using Adgf-ARNAi(Hml-gal4 UAS-Adgf-ARNAi). (G) Overexpression of Adgf-A can suppress the proliferation phenotype due to loss of
PvrRNAi(Hml-gal4 UAS-PvrRNAiUAS-Adgf-A; 17 ± 7 cells, n = 8, p < 0.03 when compared with (E).
1590 Cell 147, 1589–1600, December 23, 2011 ª2011 Elsevier Inc.
similar to vertebrate adenosine deaminases (ADAs). Adgf-A is
a secreted enzyme that converts extracellular adenosine into in-
osine by deamination (Dolezal et al., 2003; Maier et al., 2001).
Two distinct adenosine deaminases, ADA1 and ADA2/CECR1,
are found in humans. CECR1 is secreted by monocytes as
they differentiate into macrophages (Zavialov et al., 2010). In
Drosophila, mutation of Adgf-A causes increased adenosine
levels and increase in circulating blood cells (Dolezal et al.,
2005; Zurovec et al., 2002).
Extracellular adenosine is sensed by the single Drosophila
adenosine receptor (AdoR) that generates a mitogenic signal
through the G protein/adenylate cyclase/cAMP-dependent
Protein Kinase A (PKA) pathway (Dolezelova et al., 2007). A
the Hedgehog signal. We were intrigued by the potential link
between adenosine and Hedgehog signaling, both through
PKA mediated regulation of Ci, and propose a model that the
niche signal and the CZ signal interact to maintain the progenitor
population in a quiescent and undifferentiated state within the
MZ of the lymph gland.
Signals from Differentiating Hemocytes Regulate
Hematopoietic Progenitor Quiescence
During the mid second instar, blood cells initiate differentiation in
the larval lymph gland marking the beginning of cortical zone
(CZ) formation (Figure 1A). The first cells that express differenti-
ation markers appear stereotypically at the peripheral edge of
the lymph gland (Figure 1A). These differentiating cells will even-
tually populate an entire peripheral compartment that will com-
prise the CZ. The timing of the first signs of differentiation
matches closely with the onset of quiescence among the
precursor population, eventually giving rise to the medullary
The close temporal synchronization of CZ formation and the
quiescence of MZ progenitors raised the intriguing possibility
that the onset of differentiation might regulate the proliferation
profile of the progenitors. To test this hypothesis, we induced
cell death by expressing the pro-apoptotic proteins Hid and
Reaper in the differentiating hemocytes and assayed for the
effect of their loss on the progenitor population. We found that
loss of CZ cells induces proliferation of the adjacent progenitor
cells, which are normally quiescent at this stage (Figures 1B
and 1C and Figures S1A and S1B, available online).
We knocked down candidate ligands in the lymph gland by
RNA interference (RNAi) and monitored for a loss of progenitor
quiescence. This survey identified Pvf1 as a signaling molecule
that is required for the maintenance of quiescence within the
lymph gland. Expressing Pvf1RNAiusing Gal4 drivers specific to
either niche (PSC) cells using Antp-gal4, progenitor cells using
dome-gal4, or differentiating cells using Hml-gal4 showed that
eration (Figure 1D and Figure S1C), whereas Pvf1 knockdown in
progenitors or differentiating cells has no effect on the lymph
sized in the PSC is required for progenitor quiescence.
To determine the site of Pvf1 function, its receptor Pvr was
knocked down in the lymph gland using a similar approach.
Interestingly, we found that PvrRNAiexpressed under the control
of drivers specific to differentiating cells (Hml-gal4 and pxn-gal4)
causes a loss of progenitor quiescence (Figure 1E and Fig-
ure S1D). The BrdU incorporating cells do not express differen-
tiation markers (Figures S1A–S1F). Thus, differentiation follows
the proliferative event. Lymph glands are not similarly affected
when Pvr function is downregulated in the progenitors them-
selves (Figure S1J). These results indicate that Pvf1 originates
in the niche and activates Pvr in maturing hemocytes, and that
this signaling system is important for the quiescence of MZ
progenitors. These results did not explain, though, how maturing
cells might signal back to the progenitors causing them to main-
Given the previously known role of Adgf-A in the control of
hemocyte number in circulation (Dolezal et al., 2005), we inves-
tigated whether this protein plays a similar role in the lymph
gland. Remarkably, downregulation of the secreted Adgf-A pro-
tein in the differentiating hemocytes of the CZ, achieved by ex-
pressing Adgf-ARNAiunder Hml-gal4 control, induces loss of
quiescence ofMZprogenitors(Figure1Fand FigureS1E),similar
to that seen with loss of Pvr in the CZ. This suggests that Adgf-A
may act as a signal originating from differentiating hemocytes
of this idea, while overexpression of Adgf-A in differentiating
hemocytes alone does not affect normal zonation, it suppresses
the induced progenitor proliferation caused by downregulation
reasons. For loss of signaling molecules (Figures 1E and 1F), it is
the break in the signaling network necessary for reducing aden-
osine that causes continued proliferation and eventual differenti-
ation. For rpr/hid (Figure 1C) the signaling cell itself has been
removed, thereby causing a lack in a backward signal. Quantita-
tive analysis of the data (from Figures 1B, 1E, and 1G, see
legend) is consistent with a role for Adgf-A downstream of Pvr
and leads us to the model shown in Figure 1H, the mechanistic
details of which are genetically dissected in this study.
(H) The color scheme of the model (right) corresponds to the zones represented in the schematic of the lymph gland (left). Maintenance of progenitor quiescence
within the MZ requires Pvf1 from the PSC, Pvr function in the CZ, and Adgf-A from the CZ functioning downstream of Pvr.
(I–N) Expression of Pvf1 and Pvr in the developing lymph gland. TOPRO-3 (in [K]–[M], blue) marks nuclei. (I) and (M) Midsecond instar lymph glands from control
animals (Hml-gal4 UAS-2xEGFP); (J–L’) and (N) Third instar lymph glands. (I) Pvf1 expression is seen as small punctae present at high levels in the PSC
(arrowhead) and dispersed in the rest of the lymph gland (arrow). (J) Pvf1 (red [J]) expression in the PSC colocalizes with Hedgehog (green [J’]). The overlap is
shown in yellow (J’’). (K) Within the lymph gland proper, Pvf1 (red punctae [K]) colocalizes with Rab11 (green punctae [K’]), a marker for transcytosis vesicles. The
overlap is shown in yellow (K’’). (L) Pvr mutant clones (nongreen) were generated within wild-type tissue (green) and stained for Pvf1 (red punctae). Pvf1 is present
in the Pvr+tissue (arrow) but is strongly reduced in the clones lacking Pvr (arrowhead). (M and N) Temporal analysis of Pvr expression. (M) The first differentiating
cells (green, Hml-gal4 UAS-2xEGFP) show strong upregulation of Pvr expression (red, inset). (N) At later stages, Pvr expression is relatively high in the CZ as
compared to the MZ, and is lacking in the PSC (arrowhead). See also Figure S1.
Cell 147, 1589–1600, December 23, 2011 ª2011 Elsevier Inc. 1591
Signaling from the Niche to the Differentiating Cells
Pvf1 is expressed in the same cells as Hedgehog within the PSC
in the primary lobe of the second instar lymph gland (Figures 1I
and 1J). Additionally, punctate dots that label with the anti-Pvf1
antibody can be seen throughout the lymph gland (Figure 1I) in
Rab11-positive recycling vesicles (Figure 1K). When Pvf1 RNAi
is targeted to the PSC, the Pvf1-positive punctae are eliminated
throughout the lymph gland, suggesting that Pvf1 expression is
limited to the PSC and the protein is transported by transcytosis
from the PSC to the rest of the cells of the lymph gland. Addition-
ally, clones of cells created within the lymph gland that lack the
Pvr receptor are greatly reduced in Pvf1-positive punctae (Fig-
ure 1L). The simplest explanation is that intercellular Pvf1 trans-
port involves receptor-mediated endocytosis and the recycling
of Pvf1 in Rab11 positive vesicles. With regard to Pvr expres-
sion, we found that it is absent from the PSC at all develop-
mental stages. Pvr is initially expressed at low levels in early
blood progenitors, but is then strongly upregulated in the first
differentiating hemocytes of the forming CZ during the second
instar (Figure 1M). By the late third instar, all cells of the lymph
gland (except PSC cells) express detectable levels of Pvr,
although CZ cells continue to be higher in their expression
The loss of the progenitor cell quiescence phenotype upon
loss of Pvf1/Pvr signaling apparent in the second instar (Figures
1D and 1E) is manifested in the third instar as a terminal pheno-
type in which all progenitor cells of the lymph gland differentiate.
Cells occupying the position of the MZ in normal lymph glands
all express differentiation markers in the mutants, leaving no
progenitors behind. This is a correlation in every genetic back-
ground where the precursor population fails to quiesce in the
second instar and continues to incorporate BrdU; the third instar
phenotype is complete differentiation of the cells that would,
in wild-type, occupy the MZ and remain undifferentiated. This
phenotype is observed when Pvf1RNAiis expressed in the PSC
(Figure 2B) or when PvrRNAiis expressed in the differentiating
hemocytes (Figure 2C). Pvr receptor activation is reduced
when Pvf1 is downregulated in the PSC (Figures S3A and S3B).
Importantly, TUNEL staining (Figures S1K–S1N) and p35 misex-
pression studies (Figures S1O and S1P) show that PvrRNAidoes
not cause an increase in cell death levels compared to wild-
type when expressed in the developing CZ. Furthermore, cells
lacking Pvr differentiate properly as assessed by retention of
their phagocytosis function (Figure 2D). The phenotype caused
by loss of Pvf1/Pvr signaling is reminiscent of the fully differen-
tiated phenotype seen upon loss of hedgehog function (Mandal
et al., 2007). However, despite the similarity in loss of function
phenotypes, Hedgehog expression in the PSC is not altered
when Pvr is removed from the CZ (Figure 2C) or Pvf1 is removed
from the PSC (Figure 2E). This suggests that the Pvf1-Pvr
signaling process operates in parallel to the niche signal
Knockdown of the second Pvr ligand Pvf2 by RNAi in the
PSC does not cause a loss of progenitor quiescence and
subsequent differentiation (Figures S2C and S2D). Rather,
Pvf2 controls Shg (E-Cadherin) expression and prevents inap-
propriate migration of the progenitors into the CZ (Figures
Signaling within the Developing CZ
VEGFR-like receptors including Pvr function in the context of
multiple effector molecules including Ras (Learte et al., 2008;
Sims et al., 2009), Jun kinase (JNK) (Bond and Foley, 2009; Ishi-
maru et al., 2004; Mathieu et al., 2007), and STAT (Fu and Zhang,
1993; Li et al., 2002). Loss of Ras function in the CZ, achieved
using a dominant negative form, does not phenocopy Pvr loss-
of-function in these cells (Figure 2F). Also the bona fide marker
of JNK activation, puc-lacZ, is not active in the CZ (not shown).
In contrast, similar to Pvr, inactivation of Stat92E in the differen-
tiating cells of the CZ using either Stat92ERNAi(Figure 2G),
Stat92EDN(Figure 2H), or the STAT inhibitor, dPIAS (Figure 2I)
causes loss of MZ progenitors. Furthermore, single copy loss
of Pvr and Stat92E in the genetic combination PvrC2195/+;
Stat92E06346/+ gives a similar loss of progenitor cell phenotype
between Pvr and Stat92E suggests a close genetic relationship
between them within a developmental pathway (for example,
see Giansanti et al., 2004 and Rogge et al., 1991). In support of
STAT functioning downstream of Pvr in differentiating cells, mis-
expression of Stat92Eactis sufficient to suppress the PvrRNAi
phenotype (Figure 2L). Taken together, these data indicate that
the developmental role of Stat92E in MZ progenitor cell mainte-
nance is nonautonomous and attributable to a function of
Stat92E in the CZ.
Stat92E loss-of-function clones generated outside of the CZ
do not cause a differentiation phenotype (Figure 2M). Whole
animal mutants in which the entire lymph gland lacks Stat92E
do show differentiation of the progenitors (Krzemien et al.,
2007) likely due to loss of Stat92E function in the CZ. Further-
more, Stat92E canonically functions downstream of the Dome-
neither dome nor hopscotch loss of function causes a loss
of progenitor quiescence and ectopic differentiation (Figures
2N–2P). Domeless is not expressed in the CZ and as expected,
expression of domeDNin the CZ has no effect on progenitor
maintenance (Figures S3C and S3D). Therefore during normal
development STAT functions downstream of Pvr in the CZ and
the Dome/JAK-STAT pathway does not appear to have a role
in this process. As previously proposed, Dome/JAK-STAT
signaling in the MZ progenitors is required for their competency
for immune-related responses (Krzemien et al., 2007; Makki
et al., 2010).
Within the CZ, expression of a GFP reporter of active STAT
signaling can be seen in a few scattered cells (Figure 2Q1–9).
These cells are negative for MZ markers (Figure 2Q1), a subset
actively proliferate (Figure 2Q2), and they also label with an anti-
body (Janssens et al., 2010) against the activated form of the
Pvr receptor (Figures 2Q3and 2Q4). These cells express Hml
(Figure 2Q5) but lack all markers of terminal differentiation (Fig-
ure 2Q6–9). The activity of this STAT reporter is reduced in a
number of relevant genetic backgrounds involving components
of this model (Figures S3E–S3H). The reporter-expressing cells
are likely ‘‘intermediate progenitors’’ (Krzemien et al., 2010)
that receive an active Pvr/STAT signal and it is possible that
these cells are the first to initiate the backward signal re-
quired for the Domeless positive progenitor quiescence
1592 Cell 147, 1589–1600, December 23, 2011 ª2011 Elsevier Inc.
Figure 2. Role of STAT Downstream of Pvr
For uniformity, differentiating hemocytes are shown in
red even if they are marked with EGFP (Hml-gal4 UAS-
2xEGFP). Lymph glands shown are from wandering third-
instar larvae. TOPRO3 marks nuclei (blue).
(A) Control. Normal Hml-gal4 expression pattern (Hml,
(B) Pvf1RNAiexpression in the PSC (green; Antp-gal4 UAS-
2xEGFP UAS-Pvf1RNAi). All non-PSC cells of the lymph
gland express Peroxidasin (Pxn, red).
(C) PvrRNAiexpression in the CZ (Hml-gal4 UAS-2xEGFP
UAS-PvrRNAi) causes MZ progenitors to differentiate (Hml,
red), although Hedgehog (green) expression in the PSC
(arrowhead) remains normal.
(D and D’) Pvr mutant clones (PvrC2195/PvrC2195, nongreen
cells) differentiate normally as judged by their ability to
phagocytose FluoSphere beads (red), similar to neigh-
boring wild-type tissue (green).
(E) Hedgehog expression (red, inset) is unaffected when
Pvf1RNAiis expressed in the PSC (green; Ser-gal4 UAS-
2xEGFP UAS-Pvf1RNAi); yellow represents colocalization
of Hedgehog and Ser-gal4 expression.
(F) Expression of RasDNin differentiating hemocytes (Hml-
gal4 UAS-2xEGFP UAS-RasDN) does not affect progenitor
(G–I) Loss of STAT function in differentiating cells
causes progenitor differentiation (Hml, red). This pheno-
type can be induced by expressing either (G) Stat92ERNAi
(Hml-gal4 UAS-2xEGFP UAS-Dcr-2 UAS-Stat92ERNAi), (H)
Stat92EDN(Hml-gal4 UAS-2xEGFP UAS-Stat92EDN), or (I)
dPIAS (Hml-gal4 UAS-2xEGFP UAS-dPIAS).
(J) Combined single-copy-loss of Pvr and Stat92E
(PvrC2195/+; Stat92E06346/+). All progenitors express Pxn
(K) As a control, expression of Stat92Eactin differentiating
cells has no effect on progenitor fate (Hml-gal4 UAS-
2xEGFP UAS-Stat92EDNDCUAS-PvrRNAi) suppresses the
progenitor differentiation phenotype caused by PvrRNAi
alone (compare with [C]).
(M) Stat92E mutant clones (Stat92E06346/Stat92E06346)
within the MZ (lacking green, demarcated by white dots)
do not cause ectopic differentiation (lack of red within the
(N) Blocking Domeless function in MZ cells (FLP-out
gal4 UAS-domeDNclones, green, see methods for details,
demarcated by white dots), does not cause them to
differentiate (P1, red).
(O) JAK mutant animals (hopM38/hopMSV1) do not exhibit
ectopic differentiation (P1, red) of progenitors.
(P) Expression of JAKRNAispecifically in differentiating
cells (Hml-gal4 UAS-2xEGFP UAS-hopRNAi) does not
cause MZ progenitor differentiation (P1, red).
(Q1–9) In the third instar, lymph glands exhibit a small
fraction of cells that express a reporter of STAT activity
(green, 10X Stat92E-GFP). These cells are negative for
domeless (dome-MESO-lacZ [Q1]), proliferate (PH3 [Q2]),
and express Pvr (Q3) and Pvract(Q4) and Hml (Hml-ga4,
UAS-lacZ [Q5]), but lack differentiation markers: Pxn (Q6),
colocalization of the STAT reporter and Hml (anti-b-gal),
Pvract, or PH3.
(R) Schematic representation of STAT function in progenitor maintenance. Pvf1 from the PSC is transported to the differentiating cells in the CZ to activate its
Pvf1 from the PSC (in [B]) or loss of Pvr (in [C]) or STAT (in [G–I]) from the CZ results in proliferation and differentiation of MZ progenitors. See also Figure S2.
Cell 147, 1589–1600, December 23, 2011 ª2011 Elsevier Inc. 1593
A Signal from Differentiating Cells to Progenitor Cells
As alluded to in Figure 1F, the primary candidate for a secreted
signaling protein that regulates progenitor quiescence and is
derived from differentiating cells is Adgf-A (Novakova and Dole-
zal, 2011). When Adgf-A is eliminated from differentiating cells,
the progenitor population fails to quiesce, suggesting a central
role for Adgf-A in CZ-derived signaling. Ultimately, the loss of
quiescence associated with loss of Adgf-A function, either in
whole-animal mutants (Figure 3B) or when Adgf-ARNAiis ex-
pressed in differentiating cells (Figure 3C), causes complete
differentiation of the hematopoietic progenitor cell population.
The loss of progenitors is not due to defects in PSC-derived
S3J–S3L). Also, the loss of progenitors in Adgf-ARNAiback-
grounds is not associated with changes in reactive oxygen
Figure 3. Adgf-A Functions Downstream of STAT
in Hematopoietic Progenitor Maintenance
Third instar lymph glands are shown. Differentiating
hemocytes are marked in red, with nuclei marked with
TOPRO3 (blue). For uniformity, hemocytes in (A), (C), (F),
and (G) are shown in red although they are genotypically
(A) Normal expression pattern of Hml-gal4, showing
differentiating cells (Hml, red). The MZ cells lack markers
of differentiation and are marked by TOPRO3 (blue).
(B) Adgf-Akarel/Adgf-Akarelmutant lymph gland. All pro-
genitors differentiate and express Pxn (red, compare
pattern with [A]).
(C) Expression of Adgf-ARNAiin differentiating hemocytes
(Hml-gal4 UAS-2xEGFP UAS-Adgf-ARNAi) causes differ-
entiation of progenitors (Hml, red; compare with [A]).
(D) Combined single-copy-loss of Pvr and Adgf-A
(PvrC2195/+; Adgf-Akarel/+). All progenitors differentiate and
(E) Combined single-copy-loss of STAT and Adgf-A
(Stat92E06346/+ Adgf-Akarel/+). All progenitors differentiate
and express Pxn.
(F) Coexpression of Adgf-A and PvrRNAi(Hml-gal4 UAS-
progenitor differentiation phenotype caused by PvrRNAi
(compare with Figure 2C).
(G) Coexpression of Adgf-A and STATDN(Hml-gal4 UAS-
UAS-Adgf-A) suppresses the
progenitor differentiation phenotype caused by STATDN
(compare with Figure 2H).
(H) Schematic representation of Adgf-A function in
progenitor maintenance. Expression of Adgf-A from the
differentiating cells, mediated by Pvf1/Pvr and STAT
signaling, functions as the CZ signal. Therefore, loss of
Adgf-A in the CZ (as in [C]) causes progenitor differentia-
tion, and overexpression of Adgf-A in the CZ (in [F] and [G])
suppresses mutant effects of Pvr and STAT. The earlier
steps are as described in Figure 2R. See also Figure S3.
species levels, as they remain unchanged when
tiating hemocytes (Figures S3M and S3N). The
phenotypic attributes due to loss of Adgf-A are
the same as those associated with loss of Pvr
ing that Adgf-A may function in the same sig-
demonstrate that the combination of a single loss-of-function
allele of Adgf-A with a single loss-of-function allele of Pvr (Fig-
ure 3D and Figure S3I) or Stat92E (Figure 3E and Figure S3I) is
sufficient to cause a loss of progenitors in the developing lymph
gland. These results are strongly indicative of a close functional
relationship between Pvr, STAT, and Adgf-A signaling within the
developing CZ. Additionally, the phenotypes caused by loss of
Pvr or Stat92E in the differentiating cells can be suppressed by
3F and 3G), further demonstrating that Adgf-A functions down-
stream of Pvr and STAT to regulate progenitor maintenance.
we propose the signaling pathway schematized in Figure 3H.
1594 Cell 147, 1589–1600, December 23, 2011 ª2011 Elsevier Inc.
As described previously, the primary function of Adgf-A is to
inactivate adenosine, and loss of Adgf-A causes an increase in
extracellular adenosine levels (Dolezal et al., 2005; Zurovec
et al., 2002). Additionally, adenosine levels can also be regulated
by intracellular uptake mediated by the nucleoside transporter
lularly can bind the adenosine receptor AdoR to transmit a G
protein/adenylate cyclase/PKA-mediated signal into the cell.
Within the developing lymph gland, expression of ENT3RNAiin
ably because loss of ENT3 among progenitors causes a local
tiation phenotype does not occur when ENT3 is downregulated
in differentiating cells (Figure 4C). In contrast, expression of
AdoRRNAiin the progenitor population causes an expansion of
these cells at the expense of differentiation (Figure 4D) and over-
expression of AdoR in progenitors causes a loss of progenitor
maintenance (Figure 4E). These results indicate that adenosine
signaling through AdoR in progenitor cells negatively impacts
their maintenance, and that this signaling is critically controlled
during development to maintain progenitors by the expression
of Adgf-A in the newly differentiating cells.
AdoR is a seven transmembrane domain receptor that signals
through G-proteins to activate adenylate cyclase (Dolezelova
et al., 2007). Consistent with this notion, overexpression of an
inhibitory G protein, Galphai(Yu et al., 2005) in progenitors blocks
their differentiation (Figure 4F). Similarly, reduction of Drosophila
adenylate cyclase function in progenitors using rutabagaRNAi
(Figure 4G) or rutabaga mutants (Figure S5D) causes the larger
number of cells to be maintained as progenitors when compared
with wild-type (Figure S2C), suggesting that cAMP signaling nor-
mally promotes differentiation. Overexpresssion of Adgf-A in the
differentiating hemocytes does not affect the progenitor popula-
tion (Figures S3O and S3P). Finally, the phenotype associated
with loss of Pvr (which in our model will block Adgf-A function
and increase extracellular adenosine; Figures 4H and 4J) or
Adgf-A can be suppressed in an AdoR mutant background (Fig-
ure S4). In both cases, the loss of quiescence correlates with
differentiation of cells that normally occupy the MZ compart-
ment. These results further establishes a role for Pvr and Adgf-A
in counteracting adenosine signaling through AdoR and adeny-
late cyclase in progenitor cells leading to their maintenance as
quiescent hematopoietic progenitors.
A Feedback Loop that Limits Adenosine Levels
Expressing AdoRRNAi(Figure 4K) or PKARNAi(Figure 4L) in the CZ
causes nonautonomous differentiation of MZ progenitors. This
phenotype is suppressed by overexpression of Adgf-A in the CZ
Balancing PSC versus CZ Signal
PKA occupies a unique position in being associated with both
adenosine signaling (Dolezelova et al., 2007) and Hh signaling
(Collier et al., 2004). Adenosine activates and Hh inhibits PKA,
which promotes the formation of the repressor form of Ci, Ci75
(Collier et al., 2004), suggesting perhaps a link between the CZ
and the niche-derived signals. PKA function was reduced in
the MZ by expressing either the dominant negative PKAmR* or
the dominant negative regulatory subunit PKA-R2 or PKARNAi
(Figures 5A–5D). In each case the result is a normal sized lymph
gland in which there is a relative increase in the size of the MZ at
the expense of the CZ. The proliferative profile of progenitors
with loss of function PKA does not change (Figures S5A
in a progenitor state but does not cause an overall expansion in
the total number of cells in the lymph gland. Similarly, overacti-
vating PKA using a mutated mouse PKA catalytic subunit
(PKAmC*) that is resistant to inhibition by its regulatory subunit
causes differentiation of hematopoietic progenitors (Figure 5E).
Most importantly, a single copy loss of PKA in an Adgf-A mutant
background rescues the progenitor cell phenotype (Figures
5F–5H). Adenosine levels are expected to be high in an Adgf-A
mutant background, thus causing proliferation followed by
differentiation. However, simultaneously reducing PKA function
downstream of the adenosine receptor has an attenuating effect
on this signal and hence suppresses the mutant phenotype. It is
important to point out that the nature of the signal that initiates
differentiation of the precursors remains unknown and is still
operational in all of the above genetic backgrounds.
A direct readout of Hedgehog signaling is the expression
level of the active form of Ci (Ci155) and the transcriptional
target ptc. As expected, overexpression of PKA-R2, which
inhibits PKA function, causes higher levels of active Ci and
Ptc (Figure 5I and Figure S5C), consistent with increased
Hedgehog signaling. Additionally, staining for Ciactshows that
compared with wild-type, the Adgf-A/Adgf-A mutant shows
lower Ci expression; this is enhanced upon PKA/+ heterozy-
gosity (Figures S5G and S5H). The critical issue is to show
whether Ci modulation is AdoR-dependent and we tested this
possibility by decreasing the level of AdoR in MZ cells. This
causes an increase in the level of activated Ci in the progenitor
cell population showing that AdoR-mediated cAMP-dependent
PKA activation modulates Ci activity (Figure 5J and Figures S5E
and S5F). Ultimately, the Adgf-A-related signal limits the
amount of PKA activated by adenosine signaling and therefore
the degradation of active Ci. Similarly, active Hedgehog niche
signaling also inhibits PKA, limiting degradation of Ci into its
repressive form. Together, the CZ and the PSC signals maintain
a balance of Ci activity within the MZ thereby controlling quies-
cence within the hematopoietic progenitor cell population
The role of a niche signal is well established in many develop-
mental systems that involve stem cell/progenitor populations
(Fuchs et al., 2004; Losick et al., 2011; Orkin and Zon, 2008;
Scadden, 2006). In the Drosophila lymph gland the niche ex-
presses Hh and maintains a group of progenitor cells (Mandal
et al., 2007). This current study establishes an additional
mechanism, parallel to the niche signal that originates from
Cell 147, 1589–1600, December 23, 2011 ª2011 Elsevier Inc. 1595
Figure 4. Adenosine Signaling Regulates Hematopoietic Progenitor Quiescence
In (A), (B), and (D)–(G), the progenitor cell population is marked with dome-gal4, UAS-2xEGFP (green). In (A)–(N), differentiating hemocytes are shown in red. All
lymph glands shown are from wandering third instar larvae.
(A–G) Role of the adenosine transporter ENT3, the receptor AdoR, downstream signaling component Ga, and the adenylate cyclase Rutabaga. (A) Control lymph
gland showing the normal expression pattern of dome-gal4 UAS-2xEGFP (green) in progenitors and P1 (red) in differentiating cells. (B) ENT3RNAiexpression in
progenitors (dome-gal4 UAS-2xEGFP UAS-ENT3RNAi) causes a reduction in MZ size (compare with [A]) and a corresponding increase in the CZ. (C) ENT3RNAi
expression in differentiating cells (Hml-gal4 UAS-2xEGFP UAS-ENT3RNAi) does not cause progenitor differentiation. (D) AdoRRNAiexpression in progenitors
(dome-gal4 UAS-2xEGFP UAS-AdoR) causes their differentiation (compare with [A]). (F) Gaioverexpression in progenitor cells (dome-gal4 UAS-2xEGFP
UAS-Gai) blocks their differentiation. (G) RutRNAiexpression in progenitors (dome-gal4 UAS-2xEGFP UAS-RutRNAi) causes their expansion and reduces differ-
(H–J) AdoR function downstream of Pvr. (H) Control AdoR mutants (AdoRKGex/AdoRKGex) develop a small lymph gland but with normal zonation. (I) PvrRNAi
expression in differentiating cells (Hml-gal4 UAS-2xEGFP UAS-PvrRNAi) causes loss of progenitors. (J) Mutation in AdoR (AdoRKGex/AdoRKGex) partially
suppresses the loss of progenitors caused by PvrRNAiin differentiating cells (Hml-gal4 UAS-2xEGFP UAS-PvrRNAi, compare with Figure 2(C).
(K–N) AdoR and Pvr together control expression of Adgf-A in differentiating cells. (K) AdoRRNAiexpression in differentiating cells (Hml-gal4 UAS-2xEGFP UAS-
AdoRRNAi) causes progenitors to differentiate (red). (L) PKARNAiexpression in differentiating cells (Hml-gal4 UAS-2xEGFP UAS-PKARNAi) causes progenitors to
differentiate (red). (M) Coexpression of Adgf-A with AdoRRNAiin differentiating cells (Hml-gal4 UAS-2xEGFP UAS-AdoRRNAiUAS-Adgf-A) suppresses the
progenitor differentiation phenotype caused by AdoRRNAialone (compare with [K]). (N) Coexpression of Adgf-A with PKARNAiin differentiating cells (Hml-
gal4 UAS-2xEGFP UAS-PKARNAiUAS-Adgf-A) suppresses the progenitor differentiation phenotype caused by PKARNAialone (compare with [L]).
(O–P) Activation of Pvr is dependent upon the presence of AdoR. (O) A wild-type (w1118) lymph gland showing the expression pattern of activated (phosphor-
ylated) Pvr (Pvract, red). The red staining on the left is in the heart (dorsal vessel, DV). (P) In AdoRKG03964ex/AdoRKG03964exlymph glands expression of Pvr (total
protein, Pvrreg, green) is normal. However, Pvractexpression (red) is entirely missing from the lymph gland (compare with red staining in [O]). The red staining on
the left is in the dorsal vessel (DV).
(Q) Adenosine signaling in the lymph gland. Adenosine signaling through AdoR in progenitors promotes their proliferation and differentiation. This process is kept
receptor (in [E]) causes differentiation. Also in the progenitors, loss of the adenosine receptor (in (D) or its downstream components (in [F] and [G]) causes
expansion of this compartment at the cost of the CZ. In differentiating cells, Pvr collaborates with AdoR to maintain Adgf-A expression. Therefore, loss of
adenosinereceptororitsdownstreamcomponentsintheCZ(in[K]and[L])causesprogenitordifferentiation andthisphenotypeissuppressed byoverexpression
of Adgf-A. See also Figure S4.
1596 Cell 147, 1589–1600, December 23, 2011 ª2011 Elsevier Inc.
Figure 5. Interaction between PSC and CZ Signals
In (A)–(E), progenitors are marked with dome-gal4 UAS-
2xEGFP (green). In (A)–(H), the differentiating cells are
shown in red. All lymph glands shown are from wandering
third instar larvae.
(A) Control lymph gland showing the normal expression
pattern of dome-gal4 UAS-GFP (green) in progenitors and
P1 (red) in differentiating cells.
(B–E) Role of PKA in progenitor maintenance. Loss of PKA
function in progenitors causes their expansion and a cor-
responding reduction of differentiating cells, as shown by
expressing (B) the regulatory domain of PKA (dome-
gal4 UAS-2xEGFP UAS-PKAmR*), which functions as
a sink for cAMP, (C) PKAEP2132(dome-gal4 UAS-2xEGFP
UAS-PKAEP2132) which functions as a dominant-negative,
and (D) PKARNAi(dome-gal4 UAS-2xEGFP UAS-PKARNAi).
(E) Gain of PKA function in progenitors through the
expression of the constitutively active PKA (dome-
gal4 UAS-2xEGFP UAS-PKAmC*) causes their differenti-
ation. Compare with (A).
(F–H) PKA opposes the function of Adgf-A in progenitor
maintenance. (F) Normal control PKAEP2132/+. (G) Control
Adgf-Akarel/Adgf-Akarel. Hematopoietic progenitors differ-
entiate. (H) PKAEP2132/+; Adgf-Akarel/Adgf-Akarel. The
progenitor differentiation phenotype due to loss of Adgf-A
is suppressed by single copy loss of PKA (compare
(I and J) Regulation of Ciactivatedby PKA and AdoR. (I and I’)
FLP-out cell clones (green in [I], see Experimental Proce-
dures)thatreduce PKAfunction(byexpressing PKAEP2132)
exhibit high levels of Ciactivated(Ciact) expression (shown in
grayscale [I’]). (J) FLP-out cell clones (green) that reduce
AdoR function (by expressing AdoRRNAi) also exhibit
elevated levels of Ciact(shown in grayscale) expres-
(K) The niche and CZ signals function together to regulate
the levels of Ciactnecessary for progenitor maintenance in
the MZ. Note that Ci is activated not only upon loss of PKA
(in (I) but also upon loss of the adenosine receptor (in [J])
thus creating a link between Hedgehog and adenosine
signaling. See also Figure S5.
Cell 147, 1589–1600, December 23, 2011 ª2011 Elsevier Inc. 1597
differentiating cells, which also regulates quiescence of hemato-
The cells of the lymph gland proliferate at early stages, from
embryo to mid second instar. At this stage, cells farthest from
the PSC initiate differentiation and the rest enter a quiescent
phase defining a MZ. In wild-type, the cells of the MZ remain
quiescent and in progenitor form throughout the third instar
and we show here that this process requires a combination of
the PSC and CZ signals. If either signal is removed, the progen-
itor population will eventually be lost due to differentiation. In
many different genetic backgrounds, if quiescence is lost, the
progenitor population initially continues to incorporate BrdU
during the second instar without expressing any maturation
markers (Figures S1A–S1F). The differentiation phenotype,
characterized by the expression of such markers, follows this
abnormal proliferation. The net result is that whenever the
progenitors accumulate BrdU (but not express any markers of
differentiation) in the second instar, all cells of the lymph gland
are differentiated and no MZ remains in the third instar. While
we do not know the nature of the signal that triggers hemocyte
differentiation, withdrawal of Wingless may play a role in this
process (Sinenko et al., 2009).
Our experimental analysis has demonstrated a novel role for
Pvr in maturing hemocytes and its ligand, Pvf1, in the cells of
the PSC. Pvf1 expression increases at a stage when the lymph
gland is highly proliferative. At this critical window in develop-
ment, Pvf1 originating from the PSC is transported to the differ-
entiating hemocytes, binds to its receptor Pvr, and activates
a STAT-dependent signaling cascade. At this stage, Pvf1 is
sensed by all cells but it is only in the differentiating hemocytes
that it activates Adgf-A in an AdoR/Pvr-dependent manner.
This secreted factor Adgf-A is required for regulating extracel-
lular adenosine levels. High adenosine would signal through
AdoR and PKA to inactivate Ci and reduce the effects of the
niche-derived Hedgehog signal leading to differentiation of the
progenitor cells. The function of the Adgf-A signal is to reduce
this adenosine signal and therefore reinforce the maintenance
of progenitors by the Hedgehog signal. Thus, the Adgf-A and
Hh signals work in the same direction but Adgf-A does so by
negating a proliferative signal due to adenosine. In wild-type,
equilibrium is reached through a signal that does not originate
from the niche that opposes this proliferative process. The
attractive step in this model is that the CZ and niche (in this
case Hh-dependent) signals both impinge on common down-
relative to the niche and the differentiated cells. Most impor-
tantly, this is a mechanism for maintaining quiescence within
a moderately large population of cells that is not in direct contact
with a niche. By the time the three zone PSC/MZ/CZ system is
set up in the late second instar all the cells of the MZ express
high levels of E-cadherin, become quiescent and are maintained
as progenitors and are capable of giving rise to all blood cell line-
ages (Jung et al., 2005; Krzemien et al., 2010). Under such
circumstances, the interaction between a niche-derived signal
and an equilibrium signal originating from differentiating cells
can maintain homeostatic control of the progenitor population.
Several vertebrate stemcell/progenitor scenarios suchasduring
bone morphogenesis (Mendez-Ferrer et al., 2010) and hemato-
poiesis (He et al., 2009; Orkin and Zon, 2008) or in the Drosophila
intestine (Mathur et al., 2010) have progenitors and differenti-
ating cells in close proximity that could pose an opportunity for
a similar niche and differentiating cell-derived signal interaction.
In fact, evidence for such interactions have recently been pro-
vided for vertebrate skin cells (Hsu et al., 2011).
The role of small molecules such as adenosine has not yet
been adequately addressed in vertebrate progenitor mainte-
nance. A small molecule such as extracellular adenosine is
unlikely to form a gradient over the population of cells and main-
tain such a gradient over a developmental time scale. It is much
more likely that this system operates similar to the ‘‘quorum
sensing’’ mechanisms described for prokaryotes (Ng and Bass-
ler, 2009). A critical level of adenosine is required for proliferation
and by expressing the Adgf-A signal this threshold amount is
lowered, causing quiescence in the entire population.
We describe a developmental mechanism that is relevant
to the generation of an optimal number of blood cells in the
absence of any overt injury or infection. However, a system
that utilizes such a mechanism to maintain a progenitor popula-
tion could potentially sense a disruption upon induction of
various metabolic stresses to cause differentiation of myeloid
cells. Various mitochondrial and cellular stresses can cause
an increase in extracellular adenosine (Fredholm, 2007), but
whether they are relevant to this system remains to be studied.
In the past, we have observed dual use of reactive oxygen
species (ROS) as well as Hypoxia Inducible Factor-a (HIF-a) in
both development and stress response of the Drosophila hema-
topoietic system (Mukherjee et al., 2011; Owusu-Ansah and
Banerjee, 2009). Responses to injury have been described in
the Drosophila intestine (Amcheslavsky et al., 2009; Jiang
et al., 2009), and in satellite cells (Ten Broek et al., 2010) that
respond during injury, a stress related signal could be the initi-
ating factor that overrides a maintenance signal. Thus, the
rupted to promote a cellular response to stress signals.
All the RNAi stocks were obtained from VDRC (Vienna) and NIG (Kyoto), UAS-
PvrRNAi(B. Shilo), UAS-PvrDN(P. Rorth and D. Montell), 10xSTAT-GFP and
UAS-STATDN(J. Darnell), JAK alleles (M. Zeidler), Pvf2-lacZ (M.A.Yoo), UAS-
PKA mC* and mR* (J. Calderon), UAS-Gai(J.E. Hooper), UAS-Stat92Eact
and UAS-domeDN(E. Bach), Adgf-AKarel, UAS-Adgf-A, and UAS-AdoR
(T. Dolezal), and AdoRKGex(AdoRKG03964ex; A. Sehgal). All Hop alleles,
as the os1small eye phenotype is further enhanced by hypomorphic alleles
of JAK (Tsai and Sun, 2004).
Lymph glands were immunostained as previously described (Jung et al.,
2005). Antibodies: Rat aPvr and aPvf1 (B. Shilo), rabbit aHh (R. Holmgren),
rabbit aRab11 (D. Ready), mouse aPvract(P. Rorth), rat aBrdU (Sigma), rat
aCi155(DHSB), mouse aPxn (J. Fessler), and mouse aP1 (I. Ando).
For phagocytic assays, Fluorescent beads (FluoSpheres 0.1 mm carbox-
ylate-modified microsphere, Invitrogen, Carlsbad, USA) were injected into
the posterior lateral region of third instar larvae. One hour postinjection lymph
glands were dissected and processed for imaging. Progenitor and differenti-
ated cell types for all genotypes are assessed by combining 3-5 middle optical
sections of the lymph gland lobe.
1598 Cell 147, 1589–1600, December 23, 2011 ª2011 Elsevier Inc.
Flp-out clones within MZ were generated using a lineage tracing system
controlled by hand-gal4 and HmlD-gal4 as described (Evans et al., 2009).
Genotypes for, hs-flp derived clones: [hs-flp Ay-gal4 UAS-GFP UAS-domeDN].
FRT clones: [hs-flp FRT40A PvrC2195FRT40A Ubi-GFP] and [hs-flp FRT82B
Lymph glands were dissected in PBS and bathed in BrdU (75 ug/ml) in PBS
for 45 min. After fixation in 4% formaldehyde for 30 min they were washed in
PBS with Triton X-100 four times for 10 min each wash. Samples were treated
in 2 N HCl for 30 min at room temperature and washed three times in
PBT. Once blocked with normal goat serum, the samples were incubated in
anti-BrdU primary antibody overnight at 4?C. After washing, samples were
incubated in secondary antibody (anti-rat Cy3) for 2 hr at room temperature,
washed, and mounted.
Lymph glands were dissected in PBS and fixed in 4% formaldehyde for 20 min
at room temperature. After permeabilization in 13 PBS + 0.4% Triton X-100
(PBT) and 100mM sodium citrate + 0.1% PBT in ice for 3 min, samples were
washed in PBS four times for 10 min each wash. Samples were transferred
to TUNEL solution (Roche Biochemicals), incubated at 37?C for 75 min, rinsed
in 0.1% PBT, and mounted in Vectashield (Vector Laboratories).
Supplemental Information includes five figures and can be found with this
article online at doi:10.1016/j.cell.2011.11.041.
We thank members of the Banerjee lab for helpful discussions. We acknowl-
edge V.D.R.C., N.I.G., D.H.S.B., B.D.S.C., E. Bach, J. Calderon, J. Darnell,
T. Dolezal, J. Hooper, D. Montell, P. Rorth, A. Sehgal, B. Shilo, M.Yoo, and
M. Zeidler for reagents. Supported by NHLBI R01 HL067395 to UB and NHLBI
K08 HL087026 to JM.
Received: December 10, 2010
Revised: April 16, 2011
Accepted: November 16, 2011
Published: December 22, 2011
Amcheslavsky, A., Jiang, J., and Ip, Y.T. (2009). Tissue Damage-Induced
Intestinal Stem Cell Division in Drosophila. Cell Stem Cell 4, 49–61.
Baldwin, S.A., Yao, S.Y., Hyde, R.J., Ng, A.M., Foppolo, S., Barnes, K., Ritzel,
M.W., Cass, C.E., and Young, J.D. (2005). Functional characterization of novel
human and mouse equilibrative nucleoside transporters (hENT3 and mENT3)
located in intracellular membranes. J. Biol. Chem. 280, 15880–15887.
Bond, D., and Foley, E. (2009). A quantitative RNAi screen for JNK modifiers
identifies Pvr as a novel regulator of Drosophila immune signaling. PLoS
Pathog. 5, e1000655.
Collier, L.S., Suyama, K., Anderson, J.H., and Scott, M.P. (2004). Drosophila
Costal1 mutations are alleles of protein kinase A that modulate Hedgehog
signaling. Genetics 167, 783–796.
Crozatier, M., Ubeda, J.M., Vincent, A., and Meister, M. (2004). Cellular
immune response to parasitization in Drosophila requires the EBF orthologue
collier. PLoS Biol. 2, e196.
Dolezal, T., Dolezelova, E., Zurovec, M., and Bryant, P.J. (2005). A role for
adenosine deaminase in Drosophila larval development. PLoS Biol. 3, e201.
Dolezal, T., Gazi, M., Zurovec, M., and Bryant, P.J. (2003). Genetic analysis of
the ADGF multigene family by homologous recombination and gene conver-
sion in Drosophila. Genetics 165, 653–666.
Dolezelova, E., Nothacker, H.P., Civelli, O., Bryant, P.J., and Zurovec, M.
(2007). A Drosophila adenosine receptor activates cAMP and calcium
signaling. Insect Biochem. Mol. Biol. 37, 318–329.
Evans, C.J., Olson, J.M., Ngo, K.T., Kim, E., Lee, N.E., Kuoy, E., Patananan,
A.N., Sitz, D., Tran, P., Do, M.-T., et al. (2009). G-TRACE: rapid Gal4-based
cell lineage analysis in Drosophila. Nat. Methods 6, 603–605.
Fredholm, B.B. (2007). Adenosine, an endogenous distress signal, modulates
tissue damage and repair. Cell Death Differ. 14, 1315–1323.
Fu, X., and Zhang, J. (1993). Transcription factor p91 interacts with the
epidermal growth factor receptor and mediates activation of the c-fos gene
promoter. Cell 74, 1135–1145.
Fuchs, E., Tumbar, T., and Guasch, G. (2004). Socializing with the neighbors:
stem cells and their niche. Cell 116, 769–778.
Garrett, R.W., and Emerson, S.G. (2009). Bone and Blood Vessels: The Hard
and the Soft of Hematopoietic Stem Cell Niches. Cell Stem Cell 4, 503–506.
Giansanti, M.G., Farkas, R.M., Bonaccorsi, S., Lindsley, D.L., Wakimoto, B.T.,
Fuller, M.T., and Gatti, M. (2004). Genetic dissection of meiotic cytokinesis in
Drosophila males. Mol. Biol. Cell 15, 2509–2522.
He, S., Nakada, D., and Morrison, S.J. (2009). Mechanisms of stem cell self-
renewal. Annu. Rev. Cell Dev. Biol. 25, 377–406.
Hsu, Y.-C., Pasolli, H.A., and Fuchs, E. (2011). Dynamics between Stem Cells,
Niche, and Progeny in the Hair Follicle. Cell 144, 92–105.
Ishimaru, S., Ueda, R., Hinohara, Y., Ohtani, M., and Hanafusa, H. (2004). PVR
plays a critical role via JNK activation in thorax closure during Drosophila
metamorphosis. EMBO J. 23, 3984–3994.
Janssens, K., Sung, H.H., and Rorth, P. (2010). Direct detection of guidance
receptor activity during border cell migration. Proc. Natl. Acad. Sci. USA
Jiang, H., Patel, P.H., Kohlmaier, A., Grenley, M.O., McEwen,D.G.,and Edgar,
B.A. (2009). Cytokine/Jak/Stat Signaling Mediates Regeneration and Homeo-
stasis in the Drosophila Midgut. Cell 137, 1343–1355.
Jude, C.D., Gaudet, J.J., Speck, N.A., and Ernst, P. (2008). Leukemia and
hematopoietic stem cells: balancing proliferation and quiescence. Cell Cycle
Jung, S.-H., Evans, C.J., Uemura, C., and Banerjee, U. (2005). The Drosophila
lymph gland as a developmental model of hematopoiesis. Development 132,
Krzemien, J., Dubois, L., Makki, R., Meister, M., Vincent, A., and Crozatier, M.
(2007). Control of blood cell homeostasis in Drosophila larvae by the posterior
signalling centre. Nature 446, 325–328.
Krzemien, J., Oyallon, J., Crozatier, M., and Vincent, A. (2010). Hematopoietic
progenitors and hemocyte lineages in the Drosophila lymph gland. Dev. Biol.
Learte, A.R., Forero, M.G., and Hidalgo, A. (2008). Gliatrophic and gliatropic
roles of PVF/PVR signaling during axon guidance. Glia 56, 164–176.
Li, L., and Clevers, H. (2010). Coexistence of Quiescent and Active Adult Stem
Cells in Mammals. Science 327, 542–545.
Li, W.X., Agaisse, H., Mathey-Prevot, B., and Perrimon, N. (2002). Differential
requirement for STAT by gain-of-function and wild-type receptor tyrosine
kinase Torso in Drosophila. Development 129, 4241–4248.
Losick, V.P., Morris, L.X., Fox, D.T., and Spradling, A. (2011). Drosophila stem
cell niches: A decade of discovery suggests a unified view of stem cell regula-
tion. Developmental Cell 21, 159–171.
Maier, S., Podemski, L., Graham, S., McDermid, H., and Locke, J. (2001).
Characterization of the adenosine deaminase-related growth factor (ADGF)
gene family in Drosophila. Gene 280, 27–36.
Makki, R., Meister, M., Pennetier, D., Ubeda, J.M., Braun, A., Daburon, V.,
Krzemien, J., Bourbon, H.M., Zhou, R., Vincent, A., et al. (2010). A short
receptor downregulates JAK/STAT signalling to control the Drosophila cellular
immune response. PLoS Biol. 8, e1000441.
Cell 147, 1589–1600, December 23, 2011 ª2011 Elsevier Inc. 1599
Mandal, L., Martinez-Agosto, J.A., Evans, C.J., Hartenstein, V., and Banerjee, Download full-text
U. (2007). A Hedgehog- and Antennapedia-dependent niche maintains
Drosophila haematopoietic precursors. Nature 446, 320–324.
Martinez-Agosto, J.A., Mikkola, H.K., Hartenstein, V., and Banerjee, U. (2007).
The hematopoieticstemcell and itsniche:acomparativeview. Genes Dev.21,
Mathieu, J., Sung, H.-H., Pugieux, C., Soetaert, J., and Rorth, P. (2007). A
Sensitized PiggyBac-Based Screen for Regulators of Border Cell Migration
in Drosophila. Genetics 176, 1579–1590.
Mathur, D., Bost, A., Driver, I., and Ohlstein, B. (2010). A transient niche regu-
lates the specification of Drosophila intestinal stem cells. Science 327,
Mendez-Ferrer, S., Michurina, T.V., Ferraro, F., Mazloom, A.R., Macarthur,
B.D., Lira, S.A., Scadden, D.T., Ma’ayan, A., Enikolopov, G.N., and Frenette,
P.S. (2010). Mesenchymal and haematopoietic stem cells form a unique
bone marrow niche. Nature 466, 829–834.
Development 137, 27–31.
Mukherjee, T., Kim, W.S., Mandal, L., and Banerjee, U. (2011). Interaction
Between Notch and Hif-1a in Development and Survival of Drosophila Blood
Cells. Science 332, 1210–1213.
Ng, W.L., and Bassler, B.L. (2009). Bacterial quorum-sensing network archi-
tectures. Annu. Rev. Genet. 43, 197–222.
Novakova, M., and Dolezal, T. (2011). Expression of Drosophila adenosine
deaminase in immune cells during inflammatory response. PLoS ONE 6,
Orkin, S.H., and Zon, L.I. (2008). Hematopoiesis: an evolving paradigm for
stem cell biology. Cell 132, 631–644.
Owusu-Ansah, E., and Banerjee, U. (2009). Reactive oxygen species prime
Drosophila haematopoietic progenitors for differentiation. Nature 461,
Rogge, R., Karlovich, C.A., and Banerjee, U. (1991). Genetic dissection of
neurodevelopmental pathway: Son of Sevenless functions downstream of
the sevenless and EGF receptor tyrosine kinases. Cell 64, 39–48.
Scadden, D.T. (2006). The stem-cell niche as an entity of action. Nature 441,
Sims, D., Duchek, P., and Baum, B. (2009). PDGF/VEGF signaling controls cell
size in Drosophila. Genome Biol. 10, R20.
Role ofWinglessSignaling inStem-likeHematopoietic Precursor Maintenance
in Drosophila. Dev. Cell 16, 756–763.
Smelkinson, M.G., and Kalderon, D. (2006). Processing of the Drosophila
hedgehog signaling effector Ci-155 to the repressor Ci-75 is mediated by
direct binding to the SCF component Slimb. Curr. Biol. 16, 110–116.
Ten Broek, R.W., Grefte, S., and Von den Hoff, J.W. (2010). Regulatory factors
and cell populations involved in skeletal muscle regeneration. J. Cell. Physiol.
Tsai, Y.C., and Sun, Y.H. (2004). Long-range effect of upd, a ligand for Jak/
STAT pathway, on cell cycle in Drosophila eye development. Genesis 39,
Tumbar, T., Guasch, G., Greco, V., Blanpain, C., Lowry, W.E., Rendl, M., and
Fuchs, E. (2004). Defining the epithelial stem cell niche in skin. Science 303,
Yu, F., Wang, H., Qian, H., Kaushik, R., Bownes, M., Yang, X., and Chia, W.
(2005). Locomotion defects, together with Pins, regulates heterotrimeric
G-protein signaling during Drosophila neuroblast asymmetric divisions. Genes
Dev. 19, 1341–1353.
Zavialov, A.V., Gracia, E., Glaichenhaus, N., Franco, R., Zavialov, A.V., and
Lauvau, G. (2010). Human adenosine deaminase 2 induces differentiation of
monocytes into macrophages and stimulates proliferation of T helper cells
and macrophages. J. Leukoc. Biol. 88, 279–290.
Zurovec, M., Dolezal, T., Gazi, M., Pavlova, E., and Bryant, P.J. (2002). Aden-
osine deaminase-related growth factors stimulate cell proliferation in
Drosophila by depleting extracellular adenosine. Proc. Natl. Acad. Sci. USA
1600 Cell 147, 1589–1600, December 23, 2011 ª2011 Elsevier Inc.