1. Developmental Biology

The vascular niche controls Drosophila hematopoiesis via fibroblast growth factor signaling

  1. Manon Destalminil-Letourneau
  2. Ismaël Morin-Poulard
  3. Yushun Tian
  4. Nathalie Vanzo
  5. Michele Crozatier  Is a corresponding author
  1. Centre de Biologie du Développement (CBD), Centre de Biologie Intégrative (CBI), Université de Toulouse, CNRS, UPS, France
https://doi.org/10.7554/eLife.64672
Share this article
Cite this article
  1. Manon Destalminil-Letourneau
  2. Ismaël Morin-Poulard
  3. Yushun Tian
  4. Nathalie Vanzo
  5. Michele Crozatier
(2021)
The vascular niche controls Drosophila hematopoiesis via fibroblast growth factor signaling
eLife 10:e64672.
https://doi.org/10.7554/eLife.64672
  2. Download BibTeX
  3. Download .RIS

Abstract

In adult mammals, hematopoiesis, the production of blood cells from hematopoietic stem and progenitor cells (HSPCs), is tightly regulated by extrinsic signals from the microenvironment called ‘niche’. Bone marrow HSPCs are heterogeneous and controlled by both endosteal and vascular niches. The Drosophila hematopoietic lymph gland is located along the cardiac tube which corresponds to the vascular system. In the lymph gland, the niche called Posterior Signaling Center controls only a subset of the heterogeneous hematopoietic progenitor population indicating that additional signals are necessary. Here we report that the vascular system acts as a second niche to control lymph gland homeostasis. The FGF ligand Branchless produced by vascular cells activates the FGF pathway in hematopoietic progenitors. By regulating intracellular calcium levels, FGF signaling maintains progenitor pools and prevents blood cell differentiation. This study reveals that two niches contribute to the control ofDrosophila blood cell homeostasis through their differential regulation of progenitors.

Introduction

In adult mammals, HSPCs in the bone marrow ensure the constant renewal of blood cells. The cellular microenvironment of HSPCs, called ‘niche’, regulates hematopoiesis under both homeostatic and immune stress conditions (Asada et al., 2017; Calvi et al., 2003; Calvi and Link, 2015; He et al., 2014; Kiel et al., 2005; Kobayashi et al., 2016; Morrison and Scadden, 2014; Zhao and Baltimore, 2015). Recent studies have revealed significant molecular and functional heterogeneity within the HSPC pool (for review Haas et al., 2018). These findings challenge the differential contribution of niche cell types to HSPC diversity. Given the high conservation of regulatory networks between insects and vertebrates, Drosophila has become an important model to study how hematopoiesis is controlled (Evans et al., 2003; Hartenstein, 2006). Insect blood cells, or hemocytes, are related to the mammalian myeloid lineage. In Drosophila, three blood cell types are produced: plasmatocytes that are macrophages involved in phagocytosis, crystal cells involved in melanisation and wound healing and lamellocytes required for encapsulation of pathogens too large to be destroyed by phagocytosis. Lamellocytes represent a cryptic cell fate since they only differentiate at the larval stage and in response to specific immune challenges such as wasp parasitism (Lemaitre and Hoffmann, 2007). The lymph gland is the larval hematopoietic organ and is composed of paired lobes, one pair of anterior lobes and several pairs of posterior lobes, aligned along the anterior part of the cardiac tube (CT) which corresponds to the vascular system (Figure 1a and Lanot et al., 2001). In third instar larvae, the anterior lobes comprise three zones: a medullary zone (MZ) containing hematopoietic progenitors, a cortical zone (CZ) composed of differentiated blood cells, and a small group of cells called the Posterior Signaling Center (PSC) (Figure 1a and Crozatier et al., 2004; Jung et al., 2005). The PSC produces a variety of signals that regulate lymph gland homeostasis (for review see Banerjee et al., 2019; Letourneau et al., 2016; Yu et al., 2018). Recently we established that cardiac cells produce the ligand Slit which, through the activation of Robo receptors in the PSC, controls the proliferation and clustering of PSC cells and in turn their function (Morin-Poulard et al., 2016). Furthermore, the MZ progenitor population is heterogeneous and a subset of progenitors called ‘core progenitors’ which express the knot/Collier (Kn/Col) and the thioester-containing protein-4 (tep4) genes is aligned along the cardiac tube and are maintained independently from the PSC (Figure 1a and Baldeosingh et al., 2018; Benmimoun et al., 2015; Oyallon et al., 2016). Altogether, these data led us to ask whether signals derived from cardiac cells are involved in the control of lymph gland homeostasis, i.e. the balance between progenitors and differentiated blood cells, independently from the PSC. To address this possibility we performed a candidate RNAi screen in cardiac cells to identify new potential signaling pathways involved in the crosstalk between the vascular and the hematopoietic organs.

Figure 1 with 1 supplement see all
Download asset Open asset
Lymph gland organization and RNAi screen results.

(a) Representation of lymph gland anterior and posterior lobes from third instar larvae. The anterior lobe is composed of progenitors (red) and core progenitors (hatched red), and the cortical zone (CZ, green). The PSC is blue and the cardiac tube (CT)/vascular system, is orange. PC corresponds to pericardial cell. (b) Summary of the screen performed by expressing RNAi in cardiac cells using the handΔ-gal4 driver. The number of genes corresponding to the different classes of phenotype is given. Subsequent panels illustrate the control and observed lymph gland defects (c, d, g, j). Anterior lobe and PSC are delimited by white and yellow dashed lines, respectively. Black-cell-GFP (BcGFP, white) labels crystal cells and Antp (black) the PSC. (c’, d’, g’, j’) BcGFP is in green; (e, h, k) PSC cell numbers; (f, i, l) Crystal cell index. (c–f) Reducing ilp6 in cardiac cells (d, d’) augments PSC cell number (e) without affecting crystal cell differentiation (f); this defines class 1. (g–i) Knocking down dachsous (ds) in cardiac cells (g, g’) decreases PSC cell number (h) and increases crystal cell index (i); this defines class 2. (j–l) Reducing pvf3 in cardiac cells (j, j’) does not modify PSC cell number (k) but increases crystal cell differentiation (l); this defines class 3. (m, n) tep4 (red) labels core progenitors. Decrease in tep4 expression is observed when pvf3 is knocked down in cardiac cells. (o) tep4 index. For all quantifications and figures, statistical analysis t-test (Mann-Whitney nonparametric test) was performed using GraphPad Prism five software. Error bars represent SEM and *p<0,1;**p<0,01; ***p<0,001; ****p<0,0001 and ns (not significant). In all confocal pictures nuclei are labeled with Topro (blue) and scale bars = 20 µm.

Figure 1—source data 1

Results of the RNAi ligand screen RNAi was expressed in cardiac cells by using the handΔ-gal4 and/or NP1029-gal4 driver.

Crystal cells were labeled by BcGFP, PSC cells were immune-stained with Antp antibody, and to visualize the core progenitors tep4 in situ hybridization was performed. In most cases, 2 RNAi lines were tested per ligand, and at least 15 lymph glands per RNAi were analyzed. Crystal cell index and PSC cell number were established. The green and red colored boxes indicate an increase and a decrease, respectively, compared to the control. Black dashes indicate that no difference was observed compared to the control. A white box indicates that this condition was not tested. Most RNAi lines that gave a modification in crystal cell index with the handΔ-gal4 driver were also analyzed with another cardiac cell driver NP1029-gal4, and proPO antibody immunostainings were performed to visualize crystal cells. Finally, for all RNAi lines that led to a defect in crystal cell differentiation with the handΔ-gal4 driver, tep4 in situ hybridizations were performed and the tep4 index was established.

https://cdn.elifesciences.org/articles/64672/elife-64672-fig1-data1-v1.xlsx
Figure 1—source data 2

Results of the RNAi ligand screen.

https://cdn.elifesciences.org/articles/64672/elife-64672-fig1-data2-v1.xlsx

Here we show that several signals produced by cardiac cells contribute to maintain lymph gland homeostasis. We investigated in more detail the role of the Fibroblast Growth Factor (FGF) ligand Branchless (Bnl). FGF signaling is conserved during evolution and is less complex in Drosophila than in humans. Ligand binding to a FGF receptor (FGFR) promotes its dimerization, which results in its tyrosine-phosphorylation, thus providing a scaffold to recruit different partners (Muha and Müller, 2013; Ornitz and Itoh, 2015). In mammals, ligand binding to the FGFR activates Ras/Raf-Mek-MAPK, PI3K/AKT, and PLCγ-Ca2+ signaling pathways (Turner and Grose, 2010).The Drosophila genome encodes two FGF receptors, Breathless (Btl) and Heartless (Htl), and three ligands, Bnl, Thisbe (Ths) and Pyramus (Pyr), (Beiman et al., 1996; Glazer and Shilo, 1991; Gryzik and Müller, 2004; Klämbt et al., 1992; Sutherland et al., 1996). Htl is activated by Ths and Pyr, while Btl is activated by Bnl. We established that Bnl is expressed in cardiac cells and signals to its receptor Breathless (Btl) expressed in progenitors. Bnl/Btl-FGF activation controls progenitor intracellular Ca2+ concentration, probably by activating Phospholipase Cγ (PLCγ) which regulates endoplasmic reticulum Ca2+ stores. Altogether, these data strongly support the conclusion that the cardiac tube plays a role similar to a niche by regulating lymph gland hematopoiesis.

Results

A cardiac screen identifies genes controlling lymph gland homeostasis

To investigate the role of cardiac cells in the control of lymph gland hematopoiesis, we performed a functional screen based on the expression, in cardiac cells, of RNAis directed against transcripts encoding known Drosophila ligands. For this we used the cardiac handΔ-gal4 driver which is expressed in cardiac cells throughout the three larval stages (Figure 1—figure supplement 1a–c’ and Monier et al., 2005; Morin-Poulard et al., 2016) to screen a collection of RNAi lines corresponding to 49 Drosophila ligands (Figure 1—source data 2). As read-outs, we analyzed blood cell differentiation with the crystal cell reporter BcGFP (Tokusumi et al., 2009), and PSC cell numbers and morphology by performing Antennapedia (Antp) immunostaining (Mandal et al., 2007). Compared to the control, 17 RNAi lines showed lymph gland homeostasis defects that were classified into three groups (Figure 1b). Class 1: Increased PSC cell numbers but no effect on crystal cell differentiation (Figure 1c–f). Two RNAis against ilp6 and spätzle4 transcripts belong to this class (source data). Class 2: Decreased PSC cell numbers and increased crystal cell differentiation (Figure 1g–i). Only one RNAi against dachsous (ds) belongs to this class (source data) Class 3: No effect on PSC cell numbers but increased crystal cell differentiation (Figure 1j–l); 14 RNAis belong to this class. The class three phenotype strongly suggested that signals from cardiac cells could control crystal cell differentiation independently from the PSC. We extended the analysis of the 14 corresponding genes by labeling the core progenitors with tep4 in situ hybridization (Krzemień et al., 2007). Reduced tep4 expression was observed for 12 RNAi treatments out of 14 (Figure 1m–o andsource data), indicating that the corresponding genes are required in cardiac cells to maintain tep4 expression in lymph gland progenitors and to prevent crystal cell differentiation. To avoid any bias due to the handΔ-gal4 driver, we also tested NP1029-gal4, an independent cardiac cell driver (Figure 1—figure supplement 1d–f’ and Monier et al., 2005; Morin-Poulard et al., 2016). Among the 14 RNAi candidates, nine gave a similar phenotype with both drivers (source data). In conclusion, our functional screen allowed us to identify nine ligands involved in communication between cardiac cells and hematopoietic progenitors to control lymph gland homeostasis.

The FGF ligand Bnl from cardiac cells controls lymph gland homeostasis

One candidate identified in our screen was Bnl. Previous studies have shown that Htl-FGF signaling is required during both early embryogenesis for lymph gland specification (Grigorian et al., 2013; Mandal et al., 2007) and in L3 larvae to control lymph gland progenitors (Dragojlovic-Munther and Martinez-Agosto, 2013). However, no role for bnl in the lymph gland has been described yet. Since bnl knock-down in cardiac cells significantly enhanced crystal cell differentiation in the lymph gland we decided to pursue an analysis of the Bnl-FGF pathway. Since Bnl is a diffusible ligand, we first documented bnl mRNA expression by in situ hybridization. bnl is expressed in cardiac and pericardial cells (Figure 2a–a’’), in agreement with previously published data (Jarecki et al., 1999). We also observed a weak bnl expression in MZ progenitors (as labeled by domeMESO >GFP in Figure 2a–a”), in differentiating hemocytes (as labeled by hml >GFP) and in a subset of crystal cells (marked by BcGFP) whereas no expression was detected in the PSC (Figure 2—figure supplement 1a–c”). In a heterozygous bnl loss-of-function mutant context where one copy of bnl (bnlP2/+) is missing, we observed an increased number of crystal cells compared to the control (Figure 2b–d). To specifically knock down bnl in cardiac cells, we expressed bnl-RNAi under the control of the cardiac tube specific driver handΔgal4. bnl loss-of-function experiments were performed from the L2 larval stage on, after the cardiac tube had formed, to avoid possible cardiac tube morphological defects (see MM and Figure 2—figure supplement 1d–e). bnl down-regulation in cardiac cells resulted in increased differentiation of both crystal cells and plasmatocytes (Figure 2e–f,i and Figure 2—figure supplement 1f–h). Increased crystal cell differentiation was also observed using another independent bnl-RNAi line (Figure 2—figure supplement 1i–k) and with the alternative NP1029-gal4 driver (Figure 2—figure supplement 1l–n). Applying bnl knockdown only after the L2 stage by using the GAL80 ts system (McGuire et al., 2004) led to a similar crystal cell differentiation defect (Figure 2—figure supplement 1o–q). We then analyzed MZ progenitors when bnl was knocked down in cardiac cells, using DomeMESO-RFP that labels all progenitors, and tep4 and Col that are expressed in the core progenitors (Krzemień et al., 2007; Oyallon et al., 2016). Compared to wild type, a reduced expression of the three markers was observed in handΔ>bnl-RNAi lymph glands (Figure 2j–r), indicating that Bnl from cardiac cells non-cell autonomously controls MZ progenitor maintenance. Altogether, these data indicate that Bnl produced in the cardiac tube acts in third instar larvae to control lymph gland homeostasis.

Figure 2 with 2 supplements see all
Download asset Open asset
Ligand Bnl is expressed in cardiac cells and controls lymph gland homeostasis.

(a) A maximum projection of 5 confocal lymph gland sections, bnl (red) is expressed in cardiac cells and MZ progenitors that express domeMESO-GFP (green). (a’, a’’) An enlarged view, bnl is red (a’) or white (a”). A white dashed line indicates the cardiac tube. * indicates a pericardiac cell. (b, c) proPO (green) labels crystal cells. bnlP2/+ heterozygous mutant lymph glands have an increased number of crystal cells (c) compared to the control (b). (e–f, g–h) Black-cell GFP (BcGFP, green) labels crystal cells. (d, i) Crystal cell index. Co-expression of bnl and bnl-RNAi in cardiac cells restores the wildtype number of crystal cells (i). (j, k) DomeMESO-RFP (red) labels MZ progenitors. Compared to the control (j) barely detectable DomeMESO-RFP levels are observed when bnl is knocked down in cardiac cells (k). (l) DomeMESO-RFP index. (m, n) tep4 labels core progenitors. Compared to the control (m) lower levels of tep4 (red) are observed when bnl is knocked down in cardiac cells (n). (o) tep4 index. (p–q) Col labels core progenitors. Compared to the control (p) lower levels of Col are observed in the core progenitors when bnl is knocked down in cardiac cells (q). (r) Col index. (s–t’) Maximum projection of 5 confocal sections of the lymph gland expressing bnl:GFP endo (green) and Col immunostaining that labels MZ progenitors (red). Compared to the control (s, s’) a decrease in bnl:GFP endo in green (t) and white (t’) is observed when bnl is knocked down in cardiac cells. Fine and thick dashed lines indicate the MZ and CT contours, respectively. (u) Bnl:GFPendo granules ratio in the MZ.

Figure 2—source data 1

Quantification of Figure 2.

https://cdn.elifesciences.org/articles/64672/elife-64672-fig2-data1-v1.xlsx

Since bnl is transcribed in MZ progenitors, though at low levels, we also analyzed its function in these cells. Reduction of bnl expression in progenitors (dome >bnl-RNAi) led to a significant increase in crystal cell differentiation as well as a decrease in tep4 expression (Figure 2—figure supplement 2a–f), indicating that Bnl produced by MZ progenitors is required to maintain their identity and to prevent their differentiation. Altogether, these data show that Bnl is produced by both MZ and cardiac cells, and that both sources are required in the control of lymph gland homeostasis.

To determine whether Bnl produced by cardiac cells contributes to the pool of Bnl present in the MZ, we analyzed endogenous Bnl distribution. We used the bnl:GFPendo knock-in allele that recapitulates bnl expression (Du et al., 2018). In agreement with in situ bnl detection, bnl:GFPendo was found in cardiac cells and in MZ progenitors (Figure 2s–s’). However, when bnl was knocked down only in the cardiac tube (handΔ>bnl-RNAi, bnl:GFPendo; Figure 2t–u) we overserved a reduction of bnl:GFPendo both in cardiac cells and MZ progenitors thus establishing that Bnl produced by cardiac cells contributes to the global MZ Bnl pool. The concomitant increased hemocyte differentiation suggests that Bnl levels contributed by the heart are required for lymph gland homoeostasis. To further support this conclusion, we overexpressed bnl only in cardiac cells (handΔ>bnl) which led to reduced crystal cell numbers (Figure 2i), and this confirms that the level of Bnl produced by cardiac cells controls hemocyte differentiation. Finally, rescue of crystal cell numbers (Figure 2g–i) and of progenitor marker expression (Figure 2—figure supplement 2g–i) was observed with a simultaneous expression of bnl and bnl-RNAi in cardiac cells (handΔ>bnl; bnl-RNAi). Altogether, these data establish that Bnl produced by cardiac cells is required for lymph gland hematopoiesis.

Since the PSC controls lymph gland cell differentiation (Benmimoun et al., 2015; Morin-Poulard et al., 2016; Oyallon et al., 2016; Tokusumi et al., 2010), we also looked at PSC cells by analyzing the expression of the PSC marker Antp (Mandal et al., 2007) when bnl was downregulated in cardiac cells. No PSC cell number or clustering defects were observed (Figure 2—figure supplement 2j–l). Hh expression in the PSC regulates progenitors and blood cell differentiation (Baldeosingh et al., 2018; Mandal et al., 2007; Tokusumi et al., 2010). The hhF4-GFP reporter transgene (Tokusumi et al., 2010) was expressed in PSC cells in the handΔ>bnl-RNAi context similar to the control (Figure 2—figure supplement 2m–o). These data strongly suggest that cardiac cell Bnl neither affects PSC cell numbers nor Hh activity, and likely acts directly on MZ progenitors to control lymph gland homeostasis. Altogether, these data indicate that although it is transcribed in many lymph gland cells, bnl expression in cardiac cells plays an essential role in the control of lymph gland homeostasis.

The FGF receptor Btl expressed in progenitors, controls lymph gland homeostasis

Bnl activates the FGF pathway by binding Btl (Kadam et al., 2009). To document endogenous Btl expression in the lymph gland, we used the btl:cherryendo knock-in allele which recapitulates Btl expression (Du et al., 2018). Strong btl:cherryendo expression was observed in cardiac cells and lower levels in MZ progenitors (as labeled by domeMESO-GFP in Figure 3a–a”). Whereas no expression was detected in PSC cells, a very faint expression occurs in a subset of crystal cells and in most differentiating blood cells (labeled by BcGFP and Hml >GFP, respectively, in Figure 3—figure supplement 1a–c”).

Figure 3 with 1 supplement see all
Download asset Open asset
Receptor Btl is expressed in hematopoietic progenitors and required to control lymph gland homeostasis.

(a) A maximum projection of 5 confocal lymph gland sections of larvae expressing btl:cherryendo (red) and domeMESO-GFP that labels MZ progenitors (green). (a’, a’’) An enlarged view, btl:cherryendo red (a’) or white (a”). Dashed lines indicate the cardiac tube contour. btl:cherryendo is expressed in cardiac cells and MZ progenitors. (b–c, e–f) Hindsight (Hnt, green) labels crystal cells. Crystal cell differentiation is increased in btldev1/+ heterozygous mutant larvae (c) compared to the control (b). (e, f) Crystal cell numbers increase when btl is knocked down in progenitors (e) and crystal cell differentiation is rescued when a constitutive activated btl receptor (btlCA) is expressed in the btl-RNAi context (f). (d, g) Crystal cell index. (h, i) DomeMESO-LacZ (red) labels MZ progenitors. Compared to the control (h) barely detectable domeMESO-LacZ levels are observed when btl is knocked down in progenitors (i). (k, l) Lower levels of tep4 (red) are observed when btl is knocked down in progenitors (l) compared to the control (k). (j, m) DomeMESO-LacZ and tep4 index, respectively. (n, p) Crystal cell numbers decrease when a dominant negative htl receptor (htlDN) is knocked down in progenitors (n) and crystal cell differentiation is increased when htlDN is co-expressed with btl-RNAi (o). (p) Crystal cell index.

Figure 3—source data 1

Quantification of Figure 3.

https://cdn.elifesciences.org/articles/64672/elife-64672-fig3-data1-v1.xlsx

To determine whether btl is required for lymph gland homeostasis, we looked at crystal cell differentiation in a heterozygous loss-of-function mutant context, where one copy of btl is mutated (btldev1/+). The resulting crystal cell index was higher than in controls (Figure 3b–d), revealing that btl controls lymph gland homeostasis. We then knocked down btl by expressing RNAi in either MZ progenitors (dome>) or cardiac cells (handΔ>). btl downregulation in cardiac cells did not significantly affect crystal cell numbers or MZ progenitors (Figure 3—figure supplement 1d–i). In contrast, knocking down btl in MZ progenitors led to increased crystal cell (Figure 3e,g) and plasmatocyte numbers (Figure 3—figure supplement 1j–l), together with a reduced expression of the two progenitor markers domeMESO-LacZ and tep4 (Figure 3h–m). We then performed rescue experiments with a constitutively active form of Btl (btlCA, Parés and Ricardo, 2016). The expression of btlCA in progenitors (dome >btlCA) led to reduced crystal cell numbers, i.e., a phenotype opposite to that of btl loss-of-function (Figure 3g). The co-expression of btlCA and btl-RNAi in progenitors (dome >btl-RNAi>btlCA; Figure 3f,g) rescued crystal cell differentiation, confirming that Btl is required in MZ progenitors. Finally, we examined whether PSC cells were affected when btl was knocked down in progenitors. No difference in PSC cell numbers or clustering was observed compared to the control (Figure 3—figure supplement 1m–o). We conclude that btl expression in MZ progenitors is required to control lymph gland homeostasis.

As opposed to Btl-FGF inhibition, Htl-FGF pathway knock-down in MZ progenitors was reported to block blood cell differentiation (Dragojlovic-Munther and Martinez-Agosto, 2013). To investigate the relationship between the two pathways in the MZ, we performed epistasis experiments. Expression of a dominant–negative form of Htl (dome >HtlDN) in progenitors led to a decrease in crystal cell differentiation, in agreement with a previous report (Dragojlovic-Munther and Martinez-Agosto, 2013). Simultaneous expression of HtlDN and btl-RNAi (dome >HtlDN > btl-RNAi) restored a wildtype number of crystal cells (Figure 3n–p). These data indicate that there is no hierarchy between Btl-FGF and Htl-FGF pathways. They also suggest that their simultaneous activity in MZ progenitors ensures a robust regulation of hemocyte differentiation.

In conclusion, downregulating Btl in MZ progenitors causes a defect in lymph gland homeostasis similar to that caused by Bnl downregulation in cardiac cells. This strongly suggests that Bnl/Btl-FGF signaling mediates inter-organ communication between MZ progenitors and the vascular system. This leads us to propose that by acting directly on MZ progenitors, the cardiac tube plays a role similar to a niche.

Bnl secreted by cardiac cells is taken up by lymph gland progenitors

Bnl originating from cardiac cells and acting on MZ progenitors raised the question of its mode of diffusion. To investigate this question, we expressed a functional GFP-tagged version of Bnl (UAS-Bnl::GFP, Lin, 2009) in cardiac cells. In addition to the expected GFP detection in these cells, discrete GFP positive cytoplasmic punctate dots/granules were detected in MZ progenitors (Figure 4a–a’’), indicating that Bnl::GFP can propagate from cardiac to lymph gland cells. Many cytoplasmic Bnl-GFP positive punctate dots in the MZ were Btl:Cherry positive (Figure 4b–b”). To further characterize these Bnl-GFP dots, we labeled recycling vesicles and late endosomes, using the ubi-Rab11-cherryFP reporter and Rab7 immunostaining, respectively. We found that many Rab11-positive and Rab7-positive vesicles co-localized with Bnl-GFP in the MZ (Figure 4c–d”). The simplest explanation is that Bnl::GFP secreted by cardiac cells is internalized by MZ progenitors, likely through receptor-mediated endocytosis.

Figure 4 with 1 supplement see all
Download asset Open asset
Ligand Bnl secreted by cardiac cells controls lymph gland crystal cell differentiation.

(a) Active bnl::GFP fusion protein is expressed in cardiac cells using handΔ-gal4 driver. Dashed lines indicate cardiac tube and the PSC is labeled by Collier (Col, red and red arrow). (a’, a’’) An enlarged view; Bnl::GFP is green (a’) or white (a”). Bnl::GFP positive granules are detected in cardiac and lymph gland cells (arrowheads). (b–b”) Enlargement of MZ area close to the cardiac tube in larvae expressing Bnl::GFP fusion protein (green) in cardiac cells (handΔ-gal4 > Bnl::GFP) and Btl:mcherryendo (red). Bnl::GFP cytoplasmic punctate dots (green in b-b’ and white in b’) co-localize with Btl:mcherryendo (yellow and arrows in b’). (c,c”) Enlargement of MZ area close to the cardiac tube in larvae expressing ubi-Rab11cherryFP (red), a marker for recycling endocytic vesicles; Bnl::GFP fusion protein (green) is expressed in cardiac cells (handΔ-gal4 > Bnl::GFP). Bnl::GFP cytoplasmic punctate dots (green in c-c’ and white in c’) co-localize with ubi-Rab11cherryFP (yellow and arrows in c’). (d–d”) Enlargement of MZ area close to the cardiac tube in larvae expressing Bnl::GFP fusion protein (green) in cardiac cells (handΔ-gal4 > Bnl::GFP) and Rab7 immunostainings (red in d, d’ and white in d’). (d–d’) Bnl::GFP cytoplasmic punctuate dots co-localize with Rab7 positive dots (yellow and arrows in d’). (e–f’) Enlargement of lymph gland cross sections extending from the cardiac tube (CT) to the cortical zone (CZ). Bnl::GFP fusion protein, expressed in cardiac cells (handΔ-gal4 > Bnl::GFP) is green (e, f) and white (e’, f’). Knocking down sar1 in cardiac cells (f, f’) leads to a decrease in Bnl::GFP cytoplasmic punctate dots compared to the control (e, e’). (g) Quantification of Bnl::GFP cytoplasmic punctate dots/granules. (h, j) BcGFP (green) labels crystal cells. Knocking down sar1 in cardiac cells (i) increases crystal cell numbers compared to the control (h). Crystal cell differentiation rescue is observed when bnl::GFP is co-expressed with sar1-RNAi (j, k). (k) Crystal cell index.

Figure 4—source data 1

Quantification of Figure 4.

https://cdn.elifesciences.org/articles/64672/elife-64672-fig4-data1-v1.xlsx

To further confirm the role of Bnl secreted by cardiac cells, we impaired endoplasmic reticulum (ER) vesicle formation by knocking down the Secretion-associated Ras-related GTPase1 (Sar1) specifically in cardiac cells. The Sar1GTPase plays a key role in the biogenesis of transport vesicles and acts by regulating vesicular trafficking (Saito et al., 2017; Yorimitsu et al., 2014). The simultaneous expression of sar1-RNAi and Bnl::GFP in cardiac cells using the handΔ-gal4 driver (handΔ>bnl::GFP >sar1 RNAi) resulted in reduced bnl::GFP cytoplasmic punctate dots in MZ progenitors compared to the control (Figure 4e–g), further confirming that cardiac cells secrete Bnl. Previous data had established that the Slit ligand, produced by cardiac cells, activates Robo receptors in the PSC and as a consequence controls PSC cell proliferation and clustering (Morin-Poulard et al., 2016). Consistent with the impairment of cardiac cell secretion when sar1 is knocked down in cardiac cells (handΔ>and NP1029>sar1 RNAi, Figure 4—figure supplement 1a–f), we also observed an increase in PSC cell numbers as well as a slight defect in their clustering, likely an effect of Slit/Robo impairment. These data indicate that sar1 knock-down in cardiac cells impairs their secretory capacity and therefore Bnl secretion.

We then analyzed the consequences on lymph gland blood cell differentiation. When sar1 was knocked down in cardiac cells using handΔ or NP1029 drivers (Figure 4h–i,k and Figure 4—figure supplement 1g–i) crystal cell numbers were higher than the control, indicating that cardiac cell secretion capacity is necessary for lymph gland hematopoiesis. To determine whether the overexpression of bnl in cardiac cells can compensate for decreased secretion, we performed rescue experiments. Crystal cell differentiation was improved when sar1-RNAi and bnl::GFP were simultaneously expressed in cardiac cells (Figure 4j,k). In conclusion, these results indicate that Bnl secreted by cardiac cells is likely taken up by MZ progenitors to activate Btl-FGF signaling, which in turn regulates lymph gland homeostasis.

FGF activation in progenitors regulates their calcium levels

The next step was to address how Bnl/Btl-FGF signaling in progenitors controls lymph gland homeostasis. Depending on the cellular context, the FGF pathway activates the MAPK or PI3K pathways, or PLCγ that controls intracellular Ca2+ levels (Ornitz and Itoh, 2015). Inactivation of MAPK or PI3K in lymph gland progenitors leads to a phenotype opposite to that of knock-down of btl in progenitors or of bnl in cardiac cells (Dragojlovic-Munther and Martinez-Agosto, 2013). This strongly suggests that Bnl/Btl-FGF signaling in progenitors does not involve MAPK or PI3K activity. It was previously reported that intracellular Ca2+ levels regulate hematopoietic progenitor maintenance: reduction of cytosolic Ca2+ in lymph gland progenitors leads to the loss of progenitor markers and to increased blood cell differentiation (Shim et al., 2013). Since Bnl/Btl-FGF knock-down and reduction of Ca2+ in progenitors induce similar lymph gland defects, we asked whether both mechanisms were functionally linked. We investigated Ca2+ levels within MZ progenitors using the Ca2+ sensor GCaMP3, which emits green fluorescence only with high Ca2+levels (Nakai et al., 2001). This sensor is expressed under the control of the dome-gal4 driver. In agreement with previous reports, high Ca2+ levels were detected in MZ progenitors (Figure 5a and Shim et al., 2013). Knocking down btl in MZ progenitors starting from L1 stage led in third instar larvae to decreased fluorescence compared to the control, indicating a reduction in Ca2+ levels (Figure 5a–c). Since no difference with control larvae could be observed at L2 stage (Figure 5—figure supplement 1a–c), we conclude that the Bnl/Btl-FGF pathway is not required for MZ progenitor specification but is required in third instar larvae to regulate Ca2+ levels. We then asked whether restoring high Ca2+ levels in progenitors could rescue the lymph gland defects due to reduced Bnl/Btl-FGF function. When free Ca2+ was high within the progenitors, such as following overexpression of Calmodulin dependent kinase II (CaMKII; dome >CaMKII) or of IP3R (Tep4 >UAS-IP3R) which controls ER-mediated Ca2+ release into the cytosol (Shim et al., 2013), a significant decrease in crystal cell numbers was observed compared to the control (Figure 5d–h and Figure 5—figure supplement 1d–h). This indicates that high Ca2+ levels in progenitors inhibit blood cell differentiation, which is in agreement with previously published data (Shim et al., 2013). Simultaneous btl reduction and Ca2+ increase in progenitors, through CaMKII or IP3R overexpression (dome >CamKII;btl-RNAi Figure 5g–h and tep4 >IP3R>btl-RNAi; Figure 5—figure supplement 1g–h), leads to a decrease in crystal cell numbers compared to the sole btl-RNAi knockdown. Overall, these data suggest that in MZ progenitors, regulation of Ca2+ levels functions downstream of the Bnl/Btl-FGF pathway. In vertebrates, FGF activation can recruit and activate PLCγ, which induces Ca2+ release from the ER (Ornitz and Itoh, 2015). small wing (sl) encodes the sole Drosophila PLCγ homolog (Thackeray et al., 1998). To investigate the role of PLCγ in the lymph gland, we analyzed crystal cell differentiation in strong hypomorphic sl2 mutants. Compared to the control, increased crystal cell differentiation was observed (as labeled by Hnt Figure 5j,m). We then addressed the role of PLCγ specifically in progenitors. Knocking down PLCγ in progenitors (dome >sl-RNAi) led to increased crystal cell differentiation compared to the control (Figure 5i,l), revealing that PLCγ in MZ progenitors regulates lymph gland hemocyte differentiation. Finally, we performed epistasis experiments to decipher the relationship between sl and the Btl-FGF pathway. When btlCA was expressed in progenitors in a sl2 mutant context (sl2; dome >bltCA; Figure 5k,n), the crystal cell index was similar to that in the sl2 mutant alone. These data establish that sl acts downstream of the Bnl/Btl-FGF pathway. Altogether, our data support the hypothesis that in MZ progenitors, Bnl/Btl-FGF signaling leads to the activation of PLCγ, which controls Ca2+ levels and in turn hemocyte differentiation.

Figure 5 with 1 supplement see all
Download asset Open asset
Btl receptor interacts genetically with CamKII to control blood cell differentiation by preventing high Ca2+ levels in progenitors.

(a, b) GCaMP3 Ca2+ sensor (dome >UAS-GCaMP3) is white. GCaMP3 intensity decreases when btl is knocked down in MZ progenitors (b) compared to the control (a). (c) Quantification of GCaMP3 intensity. (d–g, i–k) Hnt (green) labels crystal cells. Crystal cell differentiation decrease is observed when Ca2+ levels increased due to CaMKII expression in progenitors (dome >CaMKII, f) compared to the control (d). Co-expression of CaMKII and btl-RNAi in progenitors (dome >CaMKII; btl-RNAi, g) leads to a decrease in crystal cell number compared to the btl knock-down alone (e). (h, l–n) Crystal cell index. Crystal cell differentiation increase is observed in sl2 homozygous mutant larvae (j, m) and when sl is knocked down in progenitors (i, l) compared to the control (d). (k, n) No difference in crystal cell index is observed in sl2 homozygous mutant larvae and in a sl2 homozygous mutant where btlCA is expressed in MZ progenitors (sl2; dome >btlCA).

Figure 5—source data 1

Quantification of Figure 5.

https://cdn.elifesciences.org/articles/64672/elife-64672-fig5-data1-v1.xlsx

Discussion

The control of HSPCs by a specific microenvironment called ‘niche’ is established both in mammals and in Drosophila. The niche is defined by its capacity to directly regulate, through signals, stem cells and progenitors. In the mammalian bone marrow HSPCs are under the control of the endosteal and vascular niches (Asada et al., 2017; Calvi et al., 2003; Calvi and Link, 2015; He et al., 2014; Kiel et al., 2005; Morrison and Scadden, 2014). In Drosophila, lymph gland studies have so far concentrated on the PSC acting as a niche. However, a subset of lymph gland progenitors (core progenitors), which express Col and tep4 and are aligned along the cardiac tube, is maintained in the lymph gland even when the PSC function is impaired, suggesting that other signals alongside those from the PSC are required (Baldeosingh et al., 2018; Benmimoun et al., 2015; Oyallon et al., 2016). Here, we report that cardiac cells play a role similar to a niche, since they directly control core progenitor maintenance. We show that communication between the vascular system and the lymph gland involves Bnl/Btl-FGFsignaling. Bnl secreted by cardiac cells activates Bnl/Btl-FGF in progenitors, which in turn controls hemocyte homeostasis. Our data indicate that Bnl/Btl-FGF signaling regulates lymph gland homeostasis by controlling calcium levels in progenitors via PLCγ activation (Figure 6). In a previous study, we showed that signals from the cardiac tube, namely Slit, can act on the PSC, but that no cellular communication between the cardiac tube and MZ progenitors is involved (Morin-Poulard et al., 2016). Now we establish that cardiac cells regulate the extent of progenitor differentiation in the lymph gland. Therefore, two separate niches (the PSC and the cardiac tube) control lymph gland homeostasis. While the PSC acts only on a subset of MZ progenitors (Baldeosingh et al., 2018; Oyallon et al., 2016), the cardiac tube directly regulates core progenitors and in turn all MZ progenitors (Figure 6).The identification of two niches that differentially regulate lymph gland progenitors sheds further light on the parallels existing between Drosophila lymph gland and mammalian bone marrow hematopoiesis.

Figure 6
Download asset Open asset
Two niches control lymph gland homeostasis.

(a–b) Schematic representation of third instar larvae lymph gland anterior lobes. Progenitors and core progenitors are in red and hatched red, respectively. The cortical zone (CZ) is in green, the PSC and the cardiac tube (CT)/vascular system are in blue and orange, respectively. (a) In a wildtype (WT) lymph gland, under normal conditions the PSC, the first niche identified, regulates the maintenance of the progenitor pool except for core progenitors (blue arrow). Here, we show that by directly acting on core progenitors (orange arrow) the cardiac tube corresponds to a second niche present in the lymph gland. Bnl produced by cardiac cells activates its receptor Btl in progenitors. Btl-FGF activation regulates intracellular Ca2+ levels via PLCγ, and controls the maintenance of core progenitors and in turn the whole progenitor pool. (b) When bnl or btl are knocked down in cardiac cells and progenitors, respectively, an increase in blood cell differentiation in the CZ is observed at the expense of the progenitor pool.

Btl-FGF signaling regulates trachea morphogenesis, which builds the insect respiratory system (Glazer and Shilo, 1991; Klämbt et al., 1992; Muha and Müller, 2013; Sato and Kornberg, 2002; Sutherland et al., 1996). How the ligand Bnl diffuses from its source to activate Btl in neighboring cells remains a controversial issue. Studies performed on Drosophila larval Air-Sac-Primordium (ASP), using endogenous tagged versions of Bnl and Btl, brought to light a key role of long range direct cellular contacts mediated by long thin cellular extensions called cytonemes (Roy et al., 2011; Sato and Kornberg, 2002). In this process, rather than diffusing passively, Bnl produced by wing disc cells is delivered directly to ASP cells by cytonemes to activate FGF signaling (Du et al., 2018). No cytoplasmic extensions from either cardiac cells or MZ progenitors were observed so far, ruling out the delivery of Bnl from cardiac cells though long cytoplasmic extensions. Instead, both cardiac cells and MZ progenitors are embedded in a dense network of extra-cellular matrix (ECM) (Grigorian et al., 2013; Krzemień et al., 2007; Volk et al., 2014). The role of ECM components and associated cell-surface proteins, such as heparan sulfate proteoglycans (HSPGs) (Muha and Müller, 2013), in lymph gland Btl-FGF activation deserves additional investigation.

Bnl::GFP secreted by cardiac cells is detected in MZ progenitors as cytoplasmic punctate dots positive for Rab11, a marker for recycling vesicles, for Rab7, a marker for late endosomes, and for the receptor Btl. These data suggest that Bnl secreted by cardiac cells is internalized by MZ progenitors most likely through receptor-mediated endocytosis. bnl is transcribed in MZ progenitors and Bnl produced by these cells also contributes to lymph gland homeostasis. Taking into account the FGF dose-dependent response shown to operate in vertebrates (Ameri et al., 2010; Iyengar et al., 2007), several lymph gland sources of Bnl could be necessary to reach the threshold needed to fully activate the Btl-FGF pathway in lymph gland progenitors.

Interestingly, the Htl-FGF pathway is also required in MZ progenitors with a loss-of-function phenotype (Dragojlovic-Munther and Martinez-Agosto, 2013) opposite to that of to Btl-FGF inactivation. While Htl-FGF signaling acts through Ras and MAPK activation, we show here that Btl-FGF signaling controls intracellular calcium concentration in hematopoietic progenitors probably through PLCγ activation. By performing epistasis experiments, we further establish the absence of hierarchy between Htl-FGF and Btl-FGF signaling pathways in the MZ and that both pathways are required simultaneously to control lymph gland hematopoiesis. To our knowledge, this is the first example in which Btl and Htl are both expressed and required in the same cell population. We postulate that simultaneous regulation by the two pathways and a Bnl contribution by two separate tissues confers robustness to lymph gland hematopoiesis under normal developmental conditions and flexibility in response to environmental fluctuation. Since Htl and Btl inactivation leads to opposite lymph gland phenotypes, this raises the question of their respective downstream targets.

In vertebrates, FGF signaling controls both primitive and definitive hematopoiesis (Dzierzak and Bigas, 2018; Muha and Müller, 2013; Ornitz and Itoh, 2015). Additional studies indicate that FGFR1 in adult HSPCs is activated during hematopoietic recovery following injury, in order to stimulate HSPC proliferation and mobilization (Zhao et al., 2012). Furthermore, FGF2 facilitates HSPC expansion by amplifying mesenchymal stem cells, a niche cell type (Itkin et al., 2013; Itkin et al., 2012). Thus, in vertebrate adult bone marrow, the FGF pathway plays a major role in the control of hematopoiesis both under steady state conditions and in response to an immune stress. However, deciphering how FGF controls hematopoiesis in bone marrow remains an arduous task since many FGF ligands and receptors are expressed in HSPC and/or niche cells and redundancy and compensation mechanisms between different FGF members can occur (Haas et al., 2018). Given the high conservation of signaling pathways between Drosophila and mammals, the low genetic redundancy in Drosophila and the striking similarities between mammalian bone marrow and fly lymph gland, there is promise that our newly identified regulation of FGF signaling in the lymph gland will shed light on the complex regulation of FGF signaling in mammalian bone marrow.

Materials and methods

Fly strains

Request a detailed protocol

w1118 (wild type, WT), UAS-mCD8-GFP and PG125dome-gal4 (Krzemień et al., 2007), antp-gal4 (Mandal et al., 2007), handΔ-gal4 (Morin-Poulard et al., 2016) and NP1029-gal4 (Monier et al., 2005). handΔ-gal4 corresponds to the 3rd intron of hand deleted from the specific visceral mesoderm enhancer (Popichenko et al., 2007 and Laurent Perrin personal communication). Lymph gland mcd8-GFP expression patterns under handΔ-gal4 and NP1029-gal4 drivers in L1, L2, and L3 larvae are given in Figure 1—figure supplement 1. The handΔ-gal4 is expressed in all cardiac cells, whereas the NP1029-gal4 is expressed in all cardiac cells except those that are expressing seven-up (Monier et al., 2005). Strains used are BcGFP (Tokusumi et al., 2009), bnlP2 (Sutherland et al., 1996), hhF4-GFP (Tokusumi et al., 2010), domeMESO-LacZ (Krzemień et al., 2007), domeMESO-Gal4 (Louradour et al., 2017), UAS-Bnl::GFP (Lin, 2009), UAS-btlCA on II or III (Parés and Ricardo, 2016), UAS-Bnl (Jarecki et al., 1999). Ubiquitin-rab11cherryFP (Y. Bellaiche), bnl:GFPendo and btl:cherryendo knock-in alleles (Du et al., 2018). Other strains were provided by the Bloomington (BL) and the Vienna (VDRC) Drosophila RNAi stock centers: btldev1 (BL4912), Sar1-RNAi (BL 32364, Cook et al., 2017), UAS-CaMKII (BL 29662), UAS-GCaMP3 (BL32116), UAS-IP3R (BL30741), sl-RNAi (BL32385 and BL35604), sl2 (BL724). The list of RNAi lines used for the functional screen is given in Figure 1—figure supplement 1. For RNAi treatments, UAS-Dicer two was introduced and at least two RNAi lines per gene were tested. Controls correspond to Gal4 drivers crossed with w1118. In all experiments, crosses and subsequent raising of larvae until late L1/early L2 stage were performed at 22°C, before shifting larvae to 29°C until their dissection at the L3 stage. For gal80ts experiments, crosses were initially maintained at 18°C (permissive temperature) for 3 days after egg laying, and then shifted to 29°C until dissection.

RNAi screen

Request a detailed protocol

Antenapedia (Antp) immunostaining was revealed with the ABC kit from Abcam. The images were collected with a Nikon epifluorescence microscope. PSC cell numbers were counted manually using Fiji multi-point tool software. The BcGFP and anterior lobe areas were measured. Crystal cell index corresponds to BcGFP area/anterior lobe area. 2 RNAi lines per gene were tested when available, and at least 15 lymph glands were analyzed per genotype.

Generation of DomeMESO-RFP transgenic lines

Request a detailed protocol

The domeMESO sequence from pCasHs43domeMESO-lacZ (Rivas et al., 2008) was sub-cloned into pENTR Directional, following the experimental procedure of the TOPOCloning Kit from Invitrogen. The resulting plasmid was used to generate domeMESO-RFP transgenic flies using attP/attB technology (Bischof et al., 2007). The Drosophila line was created by integration at attP-68A4 (III) sites.

Antibodies and immunostaining

Request a detailed protocol

Lymph glands were dissected and processed as previously described (Krzemień et al., 2007). Antibodies used were mouse anti-Col (1/100) (Krzemień et al., 2007), chicken anti-βgal (1/1000, Abcam), rabbit anti-RFP (1/40 000, Rockland Immunochemicals), chicken anti-GFP(1/500, Abcam), mouse anti-Antp (1/100, Hybridoma Bank), mouse anti-Hnt (1/100, Hybridoma Bank); mouse anti-P1 (1/30, I. Ando, Institute of Genetics, Biological Research Center of the Hungarian Academy of Science, Szeged, Hungary), mouse anti-proPO (1/100, T.Trenczel, Justus-Liebig-University Giessen, Giessen, Germany). Secondary antibodies were Alexa Fluor-488 and −555 conjugated antibodies (1:1000, Molecular Probes) and goat anti-Chicken Alexa Fluor-488 (1/800; Molecular Probes). Nuclei were labeled with TOPRO3 (Thermo Fisher Scientific). Immunostainings were performed as previously described (Louradour et al., 2017). For detecting bnl:GFPendo and btl:cherryendo immunostainings were performed with anti-GFP and anti-RFP, respectively.

In situ hybridization

Request a detailed protocol

The protocol was as described in Oyallon et al., 2016. For fluorescent in situ hybridization we used digoxigenin-labeled tep4 and bnl probes. For revelation, samples were incubated with sheep-anti-DIG (1/1000, Roche) followed by biotinylated donkey-anti-sheep (1/500, Roche). ABC kit from Vector Laboratory was used followed by fluorescent tyramide staining (Alexa fluor 555 or 488 conjugated tyramide from Molecular Probes). The bnl probe was transcribed in vitro using T7 RNA polymerase II, from PCR-amplified DNA sequences. Pairs of primers were used and the sequence in italics corresponds to the T7 RNA-Pol II fixation site. For bnl: primer 1: GCCATGGACAACAACTTGAC/ATGAATTCTAATACGACTCACT ATAGGGCGTCGTTACGGTCCAGATTG; primer 2: GCAAGGCCAACAAGAAGAAG/ATGAATTCTAATACGACTCACTATAGGGCCTGGTCGTTATCCTGATCC.

Quantification of PSC cell numbers

Request a detailed protocol

In all experiments, genotypes were analyzed in parallel and quantified. PSC cells were counted manually using Fiji multi-point tool software. Statistical analyses (Mann–Whitney nonparametric test) were performed using GraphPad Prism five software.

Blood cell and progenitor quantification

Request a detailed protocol

Crystal cells were visualized by either BcGFP or immunostainings with antibodies against proPO or Hnt. Plasmatocytes were labeled by P1 immunostainings. DomeMESO-RFP and DomeMESO-GFP were expressed in MZ progenitors, whereas MZ core progenitors were labeled by either tep4 in situ hybridization or Col immunostainings. Optimized confocal sections were performed on Leica SPE or SP8 microscopes for 3D reconstruction. The numbers of crystal cells, plasmatocytes and progenitors stained and anterior lobe volume (in µm3) were measured using Volocity 3D Image Analysis software (PerkinElmer). Crystal cell index: (crystal cell number/anterior lobe volume)x100000; plasmatocyte and progenitor index: (plasmatocyte or progenitor volume/anterior lobe volume)x100. At least 15 anterior lobes were scored per genotype, and experiments were reproduced at least three times. Statistical analyses (Mann–Whitney nonparametric test) were performed using GraphPad Prism five software. Since the number of lymph gland differentiated blood cells fluctuates depending on the larval stage, and to limit discrepancies in all the experiments, genotypes were analyzed in parallel.

Quantification of hhF4-GFP and UAS-GCaMP3 intensity

Request a detailed protocol

Optimized confocal sections were performed on Leica SPE or SP8 microscopes for 3D reconstruction. For hhF4-GFP, the sum intensities for GFP per PSC labeled by Col and each PSC volume (in μm3) were measured using Volocity 3D Image Analysis software (PerkinElmer). The intensity of hhF4-GFP corresponds to the sum intensity of hhF4-GFP/the PSC volume. For GCaMP3, the sum intensities for GFP per lymph gland primary lobe labeled by TOPRO and each primary lobe volume (in μm3) were measured using Volocity 3D Image Analysis software. The intensity of GCaMP3 corresponds to the sum intensity of GCaMP3/per lymph gland primary lobe volume. At least 15 anterior lobes were scored per genotype, and experiments were reproduced at least three times. Statistical analyses (Mann–Whitney nonparametric test) were performed using GraphPad Prism five software.

Quantification of the diffusion in the MZ of cytoplasmic Bnl::GFP dots, in hand >Bnl::GFP and hand >Bnl::GFP >sar1 RNAi genetic contexts

Request a detailed protocol

Optimized lymph gland confocal sections were obtained with a Leica SP8 microscope for 3D reconstruction. The maximum projection of 10 slices chosen in the middle of the stack was performed. A parallelepiped with a larger corresponding to four nuclei diameter, a width of 10 confocal slices a length corresponding to the distance from the CT to the CZ was designed. Along the length, the parallelepiped was subdivided into 11 sub parallelepipeds of similar size. The number of Bnl::GFP granules per sub parallelepiped (called interval in Figure 4g legend) was counted. Spot detector plugin from ICY software (http://icy.bioimageanalysis.org/) was used to quantify the number of Bnl::GFP dots per sub parallelepiped.

Sample size

Request a detailed protocol

n corresponds to the number of anterior lobes analyzed. Figure 1: In e, for handΔ> n = 61 and n = 26 for handΔ>ilp6 RNAi. In h, for handΔ> n = 61 and n = 24 for handD >ds RNAi. In k, handΔ> n = 61 and n = 24 for handΔ>pvf3 RNAi. In f, for handΔ> n = 19 and n = 27 for handΔ>ilp6 RNAi. In i, for handΔ> n = 26 and n = 21 for handΔ>ds RNAi. In l, handΔ> n = 26 and n = 26 for handΔ>pvf3 RNAi. In o, for handΔ> n = 14 and n = 10 for handΔ>pvf3 RNAi. Figure 2: in d, n = 12 for WT and n = 10 for bnlP2/+. In i, handΔ> n = 22, handΔ>bnl-RNAi n = 13, HandΔ>bnl n = 38 and handΔ>bnl;bnl-RNAi n = 10. In l, for handΔ> n = 24 and n = 15 for handΔ>bnl-RNAi. In o, for handΔ>n = 25 and n = 28 in handΔ>bnl-RNAi. For r, handΔ> n = 36 and n = 22 for handΔ>bnl-RNAi. In u, for handΔ>n = 16 and n = 18 in handΔ>bnl-RNAi.

Figure 3a-a" domeMESO-GFP crossed with btl:mcherryendo; 3b and k: PG125dome-gal4,UAS-dcr2 crossed with w1118; c btldev1/TM6B crossed with w1118; e and i: PG125dome-gal4,UAS-dcr2 crossed with UAS-btl-RNAi; f: PG125dome-gal4,UAS-dcr2 crossed with UAS-btlCA; UAS-btl-RNAi; h: PG125dome-gal4,UAS-dcr2; DomeMESO-LacZ crossed with w1118; i: PG125dome-gal4,UAS-dcr2; DomeMESO-LacZ crossed with UAS-btl-RNAi; n: PG125dome-gal4,UAS-dcr2 crossed with UAS-htlDN; o: PG125dome-gal4,UAS-dcr2 crossed with UAS-htlDN; UAS-btl-RNAi.

In d, n = 45 for WT and n = 30 for btl dev1/+. In g, dome >n = 93, dome >btl-RNAi n = 23, dome >btlCA n = 12 and dome >btlCA;btl-RNAi n = 20. In j, n = 27 for dome> and n = 21 for dome >btl-RNAi. In m, n = 20 for dome> and n = 24 for dome >btl-RNAi. In p, n = 16 for dome>, n = 13 for dome >htlDN. n = 18 for dome >btl-RNAi and n = 29 for dome >dome > htlDN>btl-RNAi.

Figure 4a handΔ,UAS-dcr2 crossed with UAS-bnl::GFP; 4b: handΔ,UAS-dcr2; btl:mcherryendo crossed with UAS-Bnl::GFP; 4 c: handΔ,UAS-dcr2; UAS-bnl::GFP crossed with ubi-rab11::mcherry; 4d: handΔ,UAS-dcr2 crossed with UAS-bnl::GFP; 4e: handΔ,UAS-dcr2; UAS-Bnl::GFP crossed with w1118; 4 f: handΔ,UAS-dcr2; Bnl::GFP crossed with UAS-sar1-RNAi; 4 hr: handΔ,UAS-dcr2; BcGFP crossed with w1118; 4i: handΔ,UAS-dcr2; BcGFP crossed with:UAS-sar1-RNAi; 4 j: handΔ,UAS-dcr2; BcGFP crossed with UAS-sar1-RNAi; UAS-bnl::GFP.

In k: handΔ> n = 27, handΔ>sar1 RNAi n = 14, and handΔ>sar1 RNAi; Bnl//GFP n = 17.

Figure 5a PG125dome-gal4,UAS-dcr2 crossed with UAS-GCaMP3 ; 5b : PG125dome-gal4,UAS-dcr2 crossed with UAS-GCaMP3 ; UAS-btl-RNAi ; 5d : PG125dome-gal4,UAS-dcr2 crossed with w1118; 5e :PG125dome-gal4,UAS-dcr2 crossed with UAS-btl-RNAi ; 5 f :PG125dome-gal4,UAS-dcr2 crossed with UAS-CaMKII ; 5 g :PG125dome-gal4,UAS-dcr2 crossed with UAS-CaMKII ; UAS-btl-RNAi ; 5i :PG125dome-gal4,UAS-dcr2 crossed with UAS-sl-RNAi ; 5 j : sl2 ; 5 k : sl2; dome-gal4 crossed with sl2; UAS-btl-CA .

In c, n = 41 for dome> and n = 29 for dome >btl-RNAi. In h, for dome> n = 35, n = 33 for dome >btl-RNAi, n = 35 for dome >CaMKII and n = 26 for dome >CaMKII; btl-RNAi. In l, dome> n = 27 and n = 30 in dome >sl-RNAi. In m, WT n = 20 and n = 17 in sl2. In n, sl2, dome >n = 20 and n = 21 in sl2; dome >btlCA.

Replicates

Request a detailed protocol

Figure 2a-a" domeMESO-GFP crossed with w1118; Figure 2b: w1118; Figure 2c: bnlP2/TM6B crossed with w1118; Figure 2e: handΔ,UAS-dcr2; BcGFP crossed with w1118; 2 f: handΔ,UAS-dcr2; BcGFP crossed with UAS-bnl-RNAi; 2 g: handΔ,UAS-dcr2; BcGFP crossed with UAS-bnl; 2 hr: handΔ,UAS-dcr2; BcGFP crossed with UAS-bnl;UAS-bnl-RNAi; 2 j: handΔ,UAS-dcr2; domeMESO-RFP crossed with w1118; 2 k: handΔ,UAS-dcr2; domeMESO-RFP crossed with UAS-bnl-RNAi; 2 m and p: handΔ,UAS-dcr2 crossed with w1118; 2 n and q: handΔ,UAS-dcr2 crossed with UAS-bnl-RNAi. s: handΔ,UAS-dcr2; bnl:GFPendo crossed with w1118; t: handΔ,UAS-dcr2; bnl:GFPendo crossed with UAS-bnl-RNAi.

(d, i) three independent experiments were performed and quantified. One is shown. (l) two independent experiments were performed and quantified. One is shown. (o) three independent experiments were performed and quantified. One is shown. (r) two independent experiments were performed and quantified. One is shown. (u) two independent experiments were performed and quantified. One is shown. Figure 3: (d, g) three independent experiments were performed and quantified. One is shown. (j) two independent experiments were performed and quantified. One is shown. (m) three independent experiments were performed and quantified. One is shown. (p) two independent experiments were performed and quantified. One is shown Figure 4: (k) two independent experiments were performed and quantified. One is shown. (Figure 5c,h,l–n) two independent experiments were performed and quantified. One is shown.

Drosophila genetics

Request a detailed protocol

Fly crosses for each figure:

Figure 1c-c', m handΔ,UAS-dcr2; BcGFP crossed with w1118; Figure 1d,d' : handΔ,UAS-dcr2; BcGFP crossed with UAS-ilp6-RNAi; Figure 1g,g' : handΔ,UAS-dcr2; BcGFP crossed with UAS-ds-RNAi; Figure 1j,j', n : handΔ,UAS-dcr2; BcGFP crossed with UAS-pvf3-RNAi.

Figure 1—figure supplement 1a-c handΔ,UAS-dcr2 crossed with UAS-mcd8GFP; d-f:NP1029, UAS-dcr2 crossed with UAS-mcd8GFP.

Figure 2—figure supplement 2a PG125dome-gal4,UAS-dcr2 crossed with w1118; b: PG125dome-gal4,UAS-dcr2 crossed with UAS-bnl-RNAi; d: PG125dome-gal4,UAS-dcr2 crossed with w1118; e: PG125dome-gal4,UAS-dcr2 crossed with UAS-bnl-RNAi; g: handΔ,UAS-dcr2 crossed with w1118; h: handΔ,UAS-dcr2 crossed with UAS-bnl; UAS-bnl-RNAi; j: handΔ,UAS-dcr2 crossed with w1118; k: handΔ,UAS-dcr2 crossed with UAS-bnl-RNAi; m: hhF4-GFP; handΔ,UAS-dcr2 crossed with w1118; n: hhF4-GFP; handΔ,UAS-dcr2 crossed with UAS-bnl-RNAi.

Figure 3—figure supplement 1 a : BcGFP crossed with btl:cherryendo; b : hml-gal4, UAS-mcd8-GFP crossed with btl:cherryendo; c : pcol-gal4, UAS-mcd8-GFP crossed with btl:cherryendo; d : handΔ,UAS-dcr2; BcGFP crossed with w1118; e: handΔ,UAS-dcr2; BcGFP crossed with UAS-btl-RNAi ; g: handΔ,UAS-dcr2 crossed with w1118; h: handΔ,UAS-dcr2 crossed with UAS-btl-RNAi; j and m: PG125dome-gal4,UAS-dcr2 crossed with w1118; k and n: PG125dome-gal4,UAS-dcr2 crossed with UAS-btl-RNAi.

Figure 4—figure supplement 1a handΔ, UAS-dcr2 crossed with w1118; b: handΔ, UAS-dcr2 crossed with UAS-sar1-RNAi; d and g: NP1029, UAS-dcr2 crossed with w1118; e and h: handΔ, UAS-dcr2 crossed with UAS-sar1-RNAi.

Figure 5—figure supplement 1 : a : PG125dome-gal4,UAS-dcr2 crossed with UAS-GCaMP3 ; b: PG125dome-gal4,UAS-dcr2 crossed with UAS-GCaMP3 ;UAS-btl-RNAi ; d : tep4-gal4,UAS-dcr2 crossed with w1118 ; e : tep4-gal4,UAS-dcr2 crossed with UAS-btl-RNAi ; f : tep4-gal4,UAS-dcr2 crossed with UAS-IP3R; g : tep4-gal4,UAS-dcr2 crossed with UAS-IP3R ; UAS-btl-RNAi.

Sample size

Request a detailed protocol

Figure 1—figure supplement 1 At least 10 anterior lobes for each condition were analyzed. Figure 2—figure supplement 1: (h) For handΔ n = 16 and n = 17 for handΔ>bnl-RNAi. (k) for handΔ>n = 24 and n = 23 for handΔ>bnl-RNAi. (n) for NP1029 >n = 46 and n = 29 for NP1029 >bnl-RNAi. (q) for handΔ>; tub80ts n = 20 and n = 14 for handΔ>bnl-RNAi; tub80ts. Figure 2—figure supplement 2: (c) for dome >n = 24 and n = 26 for dome >bnl-RNAi. (f) dome >n = 23 and n = 24 in dome >bnl-RNAi. (i) for handΔ>n = 13, handΔ>bnl-RNAi n = 11, handΔ>bnl n = 11 and handΔ>bnl; bnl-RNAi n = 10. (l) for handΔ>n = 49 and n = 22 for handΔ>bnl RNAi. (o) for handΔ>n = 40 and n = 22 for handΔ>bnl-RNAi. Figure 3—figure supplement 1: (f) for handΔ>n = 23 and n = 17 for handΔ>btl-RNAi. (i) for handΔ>n = 50 and n = 34 for handΔ>btl-RNAi. (l) For dome >n = 26 and n = 22 for dome >btl-RNAi. (o) for dome >n = 21 and n = 23 for dome >btl-RNAi. Figure 4—figure supplement 1: (c) for handΔ>n = 37 and n = 28 for handΔ>sar1 RNAi. (f) for NP1029 >n = 20 and n = 28 for NP1029 >sar1 RNAi. (i) for NP1029 >n = 15 and n = 28 for NP1029 >sar1 RNAi. Figure 5—figure supplement 1: (c) for dome >n = 14 and for dome >btl-RNAi n = 7. (h) for tep4 >n = 14, for tep4 >btl-RNAi n = 12, for tep4 >IP3R n = 27 and for tep4 >IP3R; btl-RNAi n = 27.

Replicates

Request a detailed protocol

Figure 2—figure supplement 1a BcGFP crossed with w1118; b: hml-gal4, UAS-mcd8GFP crossed with w1118; c: pcol-gal4, UAS-mcd8-GFP crossed with w1118; d: handΔ,UAS-dcr2 crossed with UAS-mcd8-GFP; e: handΔ,UAS-dcr2 crossed UAS-mcd8-GFP; UAS-bnl-RNAi; f: handΔ,UAS-dcr2 crossed with w1118; g: handΔ,UAS-dcr2 crossed with UAS-bnl-RNAi; i: handΔ,UAS-dcr2 crossed with w1118; j: handΔ,UAS-dcr2 crossed with UAS-bnl-RNAi 34572; l: NP1029, UAS-dcr2 crossed with w1118; m: NP1029, UAS-dcr2 crossed with UAS-bnl-RNAi; o: handΔ,UAS-dcr2;tub- gal80ts crossed with w1118; p: handΔ,UAS-dcr2;tub- gal80ts crossed with UAS-bnl-RNAi.

(h)two independent experiments were performed and quantified. One is shown. (k and n) three independent experiments were performed and quantified. One is shown. (q) two independent experiments were performed and quantified. Figure 2—figure supplement 2: (c) three independent experiments were performed and quantified. One is shown. (f and i) two independent experiments were performed and quantified. One is shown. (l and o) three independent experiments were performed and quantified. One is shown. Figure 3—figure supplement 1: (f) three independent experiments were performed and quantified. One is shown. (i). two independent experiments were performed and quantified. One is shown. (l and o) three independent experiments were performed and quantified. One is shown. Figure 4—figure supplement 1: (c, f and i) two independent experiments were performed and quantified. One is shown. Figure 5—figure supplement 1: (c and h) two independent experiments were performed and quantified. One is shown.

Data availability

All data generated or analysed during this study are included in the manuscript and supporting files. Source data files have been provided for all figures.

References

  35. Thesis
    1. Lin L
    (2009)
    Clonal Analysis of Growth Behaviors During Drosophila Larval Tracheal Development In Doctoral thesis
    University of Basel.
    1. Ornitz DM
    2. Itoh N
    (2015) The fibroblast growth factor signaling pathway
    Wiley Interdisciplinary Reviews: Developmental Biology 4:215–266.
    https://doi.org/10.1002/wdev.176
    1. Thackeray JR
    2. Gaines PC
    3. Ebert P
    4. Carlson JR
    (1998)
    Small wing encodes a phospholipase C-(gamma) that acts as a negative regulator of R7 development in Drosophila
    Development 125:5033–5042.

Decision letter

  1. Bruno Lemaître
    Reviewing Editor; École Polytechnique Fédérale de Lausanne, Switzerland
  2. Utpal Banerjee
    Senior Editor; University of California, Los Angeles, United States
  3. Guy Tanentzapf
    Reviewer; University of British Columbia, Canada

In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses.

Acceptance summary:

This paper provides mechanistic insights into the role of the cardiac tube as a vascular niche controlling blood progenitor maintenance and differentiation by the Bnl-Btl arm of FGF signalling via Calcium. The findings in the manuscript are impactful and will contribute an important aspect to the field of hematopoiesis.

Decision letter after peer review:

[Editors’ note: the authors submitted for reconsideration following the decision after peer review. What follows is the decision letter after the first round of review.]

Thank you for submitting your work entitled "The vascular niche controls Drosophila hematopoiesis via Fibroblast Growth Facto signaling" for consideration by eLife. Your article has been reviewed by three peer reviewers, one of whom is a member of our Board of Reviewing Editors, and the evaluation has been overseen by a Senior Editor. The following individual involved in review of your submission has agreed to reveal their identity: Tina Mukherjee (Reviewer #2).

Our decision has been reached after consultation between the reviewers. Based on these discussions and the individual reviews below, we regret to inform you that your work will not be considered further for publication in eLife.

While the reviewers agree that this paper has its merits and deserves publication, they feel that the current manuscript has too many conceptual concerns that need clarifications which are absolutely necessary to strengthen the core idea being proposed in the study. The general consensus was to reject the manuscript but allows re-submission. This provides more time than the usual 2 months to revise the manuscript. A re-submission will be considered as new submission. Taken in consideration the concerns, you might decide to submit the manuscript to another journal. However, if you decide to resubmit, we will try to send the manuscript to the same reviewers unless you instruct us otherwise.

Reviewer #1:

In this manuscript, Destalminil-Letourneau et al., describes a signaling crosstalk between the cardiac cells and the hematopoietic progenitors that regulate Drosophila lymph gland homeostasis independently from the PSC. The authors identified through a genetic screen the Fibroblast Growth Factor (FGF) ligand Branchless (Bnl) produced by cardiac cells that maintain hematopoietic lymph gland progenitors by controlling intracellular Ca2+ concentration via PLCγ activation. This result brings advancement in the study of central hematopoiesis in the Drosophila larvae, although, it was already shown that the cardiac cells impact central hematopoiesis via the regulation of PSC morphology. While this article has some merits, I am not sure that it is strong enough to deserve publication in eLife.

First, this newly identified pathway is not integrated with the other pathways that regulate lymph gland homeostasis already described by this team and other research groups. As such, it is unclear what are the specificity of this pathway and what it brings to the process of hematopoiesis. In order to answer these questions, it would be necessary to analyse the impact of the cardiac cell-progenitor loop on hematopoiesis during an infestation.

1) A first concern is that this new pathway involving cardia cells and progenitor through Bnl and Btl is not integrated with other pathways. There is only a little insight provided on how this signaling cascade impact differentiation except for an increase of Ca2+.

2) A second concern here is that this inter-organ communication between the cardiac cells and the lymph gland has been studied under homeostatic but not immune stress conditions. Some additional experiments are needed to show how the cardiac cells might act as a niche to regulate the hematopoietic response to immune stress such as wasp parasitism.

3) The authors demonstrated the diffusion of Bnl from the cardiac cells to the lymph gland progenitors. However, the data concerning the internalization of Bnl in the lymph gland by endocytosis is not convincing. First, Figure 4 (B,B',B') is unclear. More generally, the authors should look at the colocalization of Bnl-GFP with the early endosomal Rab5 and the late endosomal Rab7 that have a key role in the transport along the endocytic pathway. In Figure 4 (A,A'): They should use an endogenously tagged version of Btl to validate further that Bbl is secreted by the cardiac cells and not only an over-expression of Bnl tagged protein.

4) A weakness of the manuscript is that we do not know the time during larval development at which this pathway plays a role.

Reviewer #2:

In this study, the authors describe the importance of cardiac/vascular cells in regulating the hematopoietic progenitor maintenance in the lymph gland. The authors propose that FGF ligand Branchless (identified as a result of a functional screen) is emanated by the vascular cells which activates FGF receptor, Breathless in the progenitor cells, necessary to maintain their homeostasis and prevent excessive differentiation. Down-stream of Btl receptor activation, the authors demonstrate the involvement of PLC-γ mediated control of progenitor calcium levels that is necessary to execute their homeostasis. Overall, the present work attempts to address the fundamental concept of understanding unidentified signal from vascular niche to control maintenance of differential population of hematopoietic progenitor cells, via utilizing a simpler and a conserved model organism Drosophila with allied similarity with the vertebrate bone marrow system. Their perusal for finding additional signals from the vascular cells in regulating hematopoiesis stems from an earlier study where PSC dependent and independent progenitors were shown to reside in the medullary zone suggesting heterogeneity in the progenitor population (Baldeosingh et al., 2018).

Although the findings seem appealing and interesting, there are certain concerns that I have regarding the main conclusions drawn in the manuscript and the experimental strategies employed to infer them. I list them point-by-point below. It is important the authors address them to prove their model and demonstrate clarity in their proposed work.

a) Specificity of Bnl source: The major concern that I have in this manuscript is regarding the specificity by which the authors prove it is the cardiac cells derived Bnl whose function is necessary for progenitor homeostasis. Given that there are two sources of Bnl production, the cardiac cells and the progenitor cells themselves, the HandΔ-gal4 line utilized in this study is confusing in accurately dissecting the source. The specificity of HandΔ-gal4, which the authors claim is a cardiac cell-specific driver, is also reported to be expressed in early LG hematopoietic compartment cells (Hand is a direct target of Tinman and GATA factors during Drosophila cardiogenesis and hematopoiesis http://dev.biologists.org/content/132/15/3525), (Pvr expression regulators in equilibrium signal control and maintenance of Drosophila blood progenitors https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4185420/), and lineage tracing with Hand-gal4 marks lymph gland progenitor cells. Given the progenitor differentiation phenotype observed upon blocking Bnl using Dome-gal4, the interpretation of the cardiac cells being the source is confusing and needs to be clarified. I understand that the authors have utilized NP1029-gal4, but again, if this has any overlapping expression in the lymph gland remains unaddressed. A lineage tracing of NP1029-gal4 to show no overlapping expression in the lymph gland is important. Secondly, both Hand and Dome are co-expressed very early (24hours AEL) in lymph gland progenitor cells, following which Hand is only restricted to cardiac and PSC cells, while Dome continues to be expressed in progenitor cells. If the loss of progenitor maintenance in HandΔ>BnlRNAi or Dome>BnlRNAi is a consequence of this early over-lapping expression needs to be tested. Temporal analysis of the lymph gland phenotype using gal80ts based experiments should help resolve this concern.

b) Temporal role of Bnl in progenitor cells: The two Bnl sources, cardiac and blood cells, loss of Bnl in either gives the same phenotype. It is important to address the temporal requirement of Branchless and the dependence or independence of one source over the other to highlight the importance of the cardiac niche in establishing progenitor homeostasis. In its current form, the manuscript fails to highlight the importance of this niche. A comparative analysis of Bnl-GFP expression during lymph gland development (from early to late3rd instar) should be undertaken to reveal its expression profile within the cardiac cells and blood progenitor cells. Secondly, changes in Bnl-GFP pattern upon expressing BnlRNAi in cardiac cells or using Dome-gal4 will hopefully address the important contribution of cardiac niche in regulating progenitor Bnl levels. Finally, with regards to the role of Bnl, the other conceptual concern that is raised is its requirement either as a progenitor development signal or that it is required post progenitor development only as a maintenance cue. Again, using gal80ts as mentioned in the previous comment, should help clarify this aspect.

c) Calcium homeostasis and FGF signaling: Although the authors show a down-regulation of GCAMP3 expression in lymph gland progenitor cells upon loss of FGF signaling, the analysis has been mostly done in the 3rd instar lymph gland when most of the tissue is differentiated. Hence it is hard to predict if the down-regulation is indeed because of loss of FGF signaling or is a consequence of progenitor differentiation and loss of these cells that maintain elevated GCAMP3 expression. Analysis of GCAMP3 levels in 2nd instar lymph glands prior to the onset of differentiation will help resolve this matter. Secondly, over-expression of PLC-γ in BtlRNAi condition may be important to prove the connection between Btl and activation of its downstream cascade linking to Calcium homeostasis more affirmatively.

d) Cardiac niche to maintain progenitor heterogeneity: The authors talk about heterogeneity in progenitors and the importance of the cardiac niche in this "Furthermore, the MZ progenitor population is heterogeneous and a subset of progenitors, called "core progenitors", which express […]these data led us to ask whether signals derived from cardiac cells were involved in the control of lymph gland homeostasis, i.e.: the balance between progenitors and differentiated blood cells, independently from the PSC". This idea doesn't seem to crystallize in the course of the findings made. The readouts for the assays done are looking at crystal cell differentiation and progenitor maintenance status to decipher FGF signaling in progenitors with ligand contribution from cardiac cells. There needs to be some way to reconcile this either experimentally (differential effects on Tep4 and Dome expression under some genetic manipulations already shown in the manuscript) or textually in the Discussion.

Reviewer #3:

In this manuscript, Destalminil-Letourneau et al. describe a novel mechanism of progenitor maintenance in the hematopoietic organ, the lymph gland. They provide evidence for the presence of a vascular niche via cardiac cells, that regulates blood progenitor maintenance in the lymph gland. This manuscript provides interesting and novel insights by showing that the existence of a vascular niche as a conserved mechanism of blood stem cell maintenance as it also exists in flies. The authors provide mechanistic insight by providing data that supports the assertion that vascular niche-lymph gland mediated FGF pathway (via Bnl-Btl) signalling positively regulates hematopoietic progenitor maintenance by controlling intracellular calcium levels in the medullary zone of the lymph gland. The manuscript reports some exciting findings but could benefit from a few improvements requiring further experimentation, validation, and analysis. Major comments are as follows:

1) In Figure 2A-A' the authors that the Bnl signal is present throughout the primary lymph gland lobe. This raises the possibility that cells in Posterior Signaling Centre or the Cortical Zone can also be a source of Bnl for the prohemocytes via a type of reciprocal signalling. One way to look at the possibility that the PSC cells or CZ cells are a potential source of Bnl is to include high resolution images where Bnl is detected (either via antibody or in situ) and simultaneously label PSC cells (with collier or Antp-Gal4 driven mCD8GFP) or CZ cells (for example with Hmldelta-Gal4 driven mCD8GFP) with an in-situ against Bnl or with Bnl antibody – to rule out the possibility.

2) The author's arguments about tissue specific requirement of Bnl would be strengthened by testing whether the increased differentiation in bnlP2 mutants can be rescued by restoring bnl levels by transgene-mediated expression in the cardiac tube and the MZ. For example it could be determined whether the constitutive activation of Btl in prohemocytes in the genetic background of Btl mutants used (btldev1/+) rescues the crystal cell and plasmatocyte differentiation. Additionally, the authors could check if expressing Bnl in the MZ using dome or tep4-Gal4 can rescue the BnlRNAi phenotype of increased prohemocyte differentiation.

3) Figure 3A also shows strong expression of Btl receptor in the cells towards the cortical zone (periphery of the LG). This raises some intriguing mechanistic questions. First, the manuscript would greatly benefit with a better, systematic, analysis of Bnl and Btl expression both in each of the zonal compartments with appropriate markers (PSC, MZ and CZ) – with high magnification/high resolution images. Second, One can envision a scenario where are the cells in the CZ themselves act as a source of Bnl that binds to Btl in neighboring (MZ) cells to regulate differentiation. Currently the authors have not ruled out this alternative model. If the Bnl is indeed secreted by differentiated cells then there could be 2 modes of signalling – either cell autonomous/paracrine mode or reciprocal signalling to maintain the MZ. It would be helpful if further clarification is provided of the role played by Bnl/Btl-FGF signalling in the CZ. It is possible is that the Bnl secreted by the cardiac cells is transcytosed/transported to the differentiated cells where it regulates differentiation cell autonomously. The authors could test this by perturbing Bnl and Btl in the CZ (using Hmldelta-Gal4 or eater-Gal4 for plasmatocytes and lz-Gal4 for crystal cells) and asking whether this affects crystal cell or plasmatocyte differentiation cell autonomously.

4) As noted by the authors it was previously shown that another FGF pathway in Drosophila, mediated by Heartless, is required for regulating hematopoiesis in the lymph gland (Dragojlovic-Munther et al., 2013). In this previous report overexpressing Btl in the progenitors had no effect (Figure S6B and K of that paper), this merits some discussion ad explanation from the authors. Furthermore, since both Btl and Heartless signalling work through the transcription effector gene pointed we would expect cross talk between the two FGF pathways. This raises an important question of how the two pathways coordinated in the lymph gland. Some genetic interaction analysis between the components of the two arms of FGF signalling will be particularly useful. Specifically, determining if one of the FGF pathways is epistatic to the other, as well as establishing whether indeed both pathways converge on Pointed.

5) The Rab11 data in Figure 4B-B' is not as strong as it could be. First, the addition of a cell membrane marker and images of a better and higher resolution would allow greater clarity. Also, since Rab11 marks the recycling endosome, it would be more appropriate to look at other endocytic markers like Rab5 (early endosome marker) and/or Rab7 (late endosome) in addition to Rab11. Also, the co-expression data in the above cases with the Rabs is currently qualitative. It is appropriate in this sort of instance to include quantitative measurement of co-localization. The authors should consider including a staining showing endocytic localization (marked with the Rabs) of the Btl receptor (by in-situ or by antibody) in the prohemocytes. Also in Figure 4B-B' shows only two cells which have Rab11-bnl co-expression/co-localization whereas the field of interest seems to have many more cells in the area. Why don't all the prohemocytes show this co-expression? Finally, also in Figure 4 Bnl diffusion data in a-a' does not look convincing without a cellular marker to identify the hemocyte populations in the LG. Such a marker should be added.

https://doi.org/10.7554/eLife.64672.sa1

Author response

[Editors’ note: the authors resubmitted a revised version of the paper for consideration. What follows is the authors’ response to the first round of review.]

Reviewer #1:

In this manuscript, Destalminil-Letourneau et al., describes a signaling crosstalk between the cardiac cells and the hematopoietic progenitors that regulate Drosophila lymph gland homeostasis independently from the PSC. The authors identified through a genetic screen the Fibroblast Growth Factor (FGF) ligand Branchless (Bnl) produced by cardiac cells that maintain hematopoietic lymph gland progenitors by controlling intracellular Ca2+ concentration via PLCγ activation. This result brings advancement in the study of central hematopoiesis in the Drosophila larvae, although, it was already shown that the cardiac cells impact central hematopoiesis via the regulation of PSC morphology. While this article has some merits, I am not sure that it is strong enough to deserve publication in eLife.

We were very surprised and disappointed by this comment, and thus we want to clarify what the novelties in this study are and their important impact for future investigations. Previous studies established that the PSC acts as a niche to control lymph gland homeostasis (for review Letourneau et al., 2016 and Banerjee et al., 2019) The hematopoietic progenitor pool in the lymph gland is heterogeneous, and a subset is not controlled by PSC signals (Baldeosingh et al., 2018). We showed previously that the cardiac/vascular system controls the PSC size and its function. Thus, through an indirect regulation involving the PSC, cardiac cells control lymph gland hematopoiesis (Morin–Poulard et al., 2016).

In this study, for the first time we provide evidence that the vascular system which directly controls blood cell progenitors independently from the PSC acts as a niche. Thus, these data establish that 2 niches control the lymph gland. Furthermore, we provide evidence that through the activation of Fibroblast Growth Factor (FGF) signaling, the vascular system prevents hematopoietic progenitors from massive differentiation, ensuring the proper balance between blood cell populations within the lymph gland. Finally, FGF activation in blood cell progenitors prevents their differentiation by regulating their intracellular calcium levels probably via PLCg activation. We sincerely think that these data bring very important and novel knowledge on the mechanisms that control lymph gland hematopoiesis and open new avenues of investigation.

First, this newly identified pathway is not integrated with the other pathways that regulate lymph gland homeostasis already described by this team and other research groups. As such, it is unclear what are the specificity of this pathway and what it brings to the process of hematopoiesis. In order to answer these questions, it would be necessary to analyse the impact of the cardiac cell-progenitor loop on hematopoiesis during an infestation.

1) A first concern is that this new pathway involving cardia cells and progenitor through Bnl and Btl is not integrated with other pathways. There is only a little insight provided on how this signaling cascade impact differentiation except for an increase of Ca2+.

We think that establishing that Bnl/Btl activation in lymph gland progenitors acts by controlling Ca2+ levels through the activation of PLCg is not of little insight but rather reveals a novel yet unsuspected mechanism. Furthermore, we analyze the relationship between BtlFGF and Htl-FGF signaling that was previously reported to be required in MZ progenitors (Dragojlovic-Munther and Martinez-Agosto, 2013). We provide evidence that there is no hierarchy between Btl-FGF and Htl-FGF signaling in the MZ. Altogether, these data reveal a primary level of integration that occurs in the lymph gland.

Unraveling the integration of Bnl/Btl activation with other pathways that regulate lymph gland homeostasis represents a huge task, since many regulators (and the list is still growing) have been identified, and there is evidence for both intrinsic and extrinsic regulations (for a review please see Letourneau et al., 2016 and Banerjee et al., 2019). If the analysis is restricted to the mechanisms that control lymph gland MZ progenitors, two main problems arise. Indeed, most studies examining the role of genes/signaling pathways involved in MZ progenitors do not distinguish whether they are required for progenitor specification at the L1 or L2 larval stages, or for progenitor maintenance at the L3 stage. In addition, most of them analyze the lymph gland progenitor pool as labelled by dome>GFP without looking at defects in the “core progenitor” pool (labelled by Col and tep4). Thus, before being able to “integrate Bnl/Btl signaling”, one would have to reinvestigate the previously identified genes/signaling pathways with respect to when and in which lymph gland progenitor pools they are required. This represents as such entirely independent studies that could take years and are out of the focus of this paper.

2) A second concern here is that this inter-organ communication between the cardiac cells and the lymph gland has been studied under homeostatic but not immune stress conditions. Some additional experiments are needed to show how the cardiac cells might act as a niche to regulate the hematopoietic response to immune stress such as wasp parasitism.

We performed wasp parasitism when bnl (hand>bnl-RNAi) or btl (dome>btl-RNAi) were knocked down in cardiac cells and progenitors respectively, and measured the % of wasp egg encapsulation. In both conditions, a defect in wasp egg encapsulation was observed compared to control, indicating that Bnl from cardiac cells and Btl in progenitors are required for an efficient cellular immune response against wasps. But now a detailed analysis is required to decipher how Btl-FGF signaling is involved. This question is out of the scope of this paper and will be the focus of intense future analyses.

3) The authors demonstrated the diffusion of Bnl from the cardiac cells to the lymph gland progenitors. However, the data concerning the internalization of Bnl in the lymph gland by endocytosis is not convincing. First, Figure 4 (B,B',B') is unclear. More generally, the authors should look at the colocalization of Bnl-GFP with the early endosomal Rab5 and the late endosomal Rab7 that have a key role in the transport along the endocytic pathway. In Figure 4 (A,A'): They should use an endogenously tagged version of Btl to validate further that Bbl is secreted by the cardiac cells and not only an over-expression of Bnl tagged protein.

In agreement with the reviewer’s comment we better defined bnl::GFP propagation from cardiac cells to MZ progenitors (HandD>Uas bnl::GFP). We used the ubi-Rab11-cherryFP reporter and performed Rab7 immunostaining, which label recycling vesicles and late endosomes, respectively. Unfortunately we cannot look at Rab5 vesicles since the Rab5 reporter is tagged by GFP and therefore colocalisation with Bnl::GFP cannot be done. We also performed Col immunostaining to visualize core progenitors. These additional data are provided in Figure 4 A, C-D”. In addition, we looked at Bnl::GFP colocalisation with the Btl receptor in progenitors (Figure 4 B-B”’) and finally at Bnl::GFP diffusion from the CT to the CZ when the secretion of cardiac cells was impaired by sar1-RNAi expression. These data are provided in Figure 4 E-G. Altogether, they strongly support the proposition that Bnl::GFP secreted by cardiac cells is internalized in MZ progenitors likely through receptor-mediated endocytosis. This is discussed in the revised version of the manuscript.

Finally we analyzed endogenous Bnl expression using the bnl:GFPendo knock-in allele (Du et al., 2018). When bnl was knocked-down in the endo cardiac tube (handD>bnl RNAi, bnl:GFPendo) lower Bnl:GFPendo levels were recorded in MZ progenitors (Figure 2S-U). This result illustrates that Bnl secreted by cardiac cells contributes to Bnl levels in progenitors. These new data have been added in the figures and in the text.

4) A weakness of the manuscript is that we do not know the time during larval development at which this pathway plays a role.

Sorry for not being clear enough in the previous version of the manuscript. In all experiments, crosses and subsequent raising of larvae until late L1/early L2 stage were performed at 22°C, before shifting larvae to 29°C until their dissection at the L3 stage. This information is now given both in the Materials and methods and the Results sections. Furthermore, according to the reviewer’s advice we performed temporal analysis of the lymph gland phenotype using tub-gal80ts (Figure 2—figure supplement O-Q) and looked at GCaMP3-expression at the L2 stage (Figure 5—figure supplement 1A-C). In conclusion, bnl/btl-FGF signaling is not involved in progenitor development till L2 but is required in third instar larvae to control the balance between MZ progenitor maintenance and hemocyte differentiation. This is now clearly stated in the revised manuscript.

Reviewer #2:

In this study, the authors describe the importance of cardiac/vascular cells in regulating the hematopoietic progenitor maintenance in the lymph gland. The authors propose that FGF ligand Branchless (identified as a result of a functional screen) is emanated by the vascular cells which activates FGF receptor, Breathless in the progenitor cells, necessary to maintain their homeostasis and prevent excessive differentiation. Down-stream of Btl receptor activation, the authors demonstrate the involvement of PLC-γ mediated control of progenitor calcium levels that is necessary to execute their homeostasis. Overall, the present work attempts to address the fundamental concept of understanding unidentified signal from vascular niche to control maintenance of differential population of hematopoietic progenitor cells, via utilizing a simpler and a conserved model organism Drosophila with allied similarity with the vertebrate bone marrow system. Their perusal for finding additional signals from the vascular cells in regulating hematopoiesis stems from an earlier study where PSC dependent and independent progenitors were shown to reside in the medullary zone suggesting heterogeneity in the progenitor population (Baldeosingh et al., 2018).

Although the findings seem appealing and interesting, there are certain concerns that I have regarding the main conclusions drawn in the manuscript and the experimental strategies employed to infer them. I list them point-by-point below. It is important the authors address them to prove their model and demonstrate clarity in their proposed work.

a) Specificity of Bnl source: The major concern that I have in this manuscript is regarding the specificity by which the authors prove it is the cardiac cells derived Bnl whose function is necessary for progenitor homeostasis. Given that there are two sources of Bnl production, the cardiac cells and the progenitor cells themselves, the HandΔ-gal4 line utilized in this study is confusing in accurately dissecting the source. The specificity of HandΔ-gal4, which the authors claim is a cardiac cell-specific driver, is also reported to be expressed in early LG hematopoietic compartment cells (Hand is a direct target of Tinman and GATA factors during Drosophila cardiogenesis and hematopoiesis http://dev.biologists.org/content/132/15/3525), (Pvr expression regulators in equilibrium signal control and maintenance of Drosophila blood progenitors https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4185420/), and lineage tracing with Hand-gal4 marks lymph gland progenitor cells. Given the progenitor differentiation phenotype observed upon blocking Bnl using Dome-gal4, the interpretation of the cardiac cells being the source is confusing and needs to be clarified. I understand that the authors have utilized NP1029-gal4, but again, if this has any overlapping expression in the lymph gland remains unaddressed. A lineage tracing of NP1029-gal4 to show no overlapping expression in the lymph gland is important. Secondly, both Hand and Dome are co-expressed very early (24hours AEL) in lymph gland progenitor cells, following which Hand is only restricted to cardiac and PSC cells, while Dome continues to be expressed in progenitor cells. If the loss of progenitor maintenance in HandΔ>BnlRNAi or Dome>BnlRNAi is a consequence of this early over-lapping expression needs to be tested. Temporal analysis of the lymph gland phenotype using gal80ts based experiments should help resolve this concern.

We apologize for creating confusion about the HandD-gal4 driver we used. The HandD-gal4 diver is NOT the hand-gal4 driver (also called hand cardiac and hematopoiesis (HCH)-gal4) described in (Han and Olson, 2005) and used in the 2 papers cited by the reviewer. The HandD-gal4 has been used previously in Morin–Poulard et al., 2016. In larvae, the HandD-gal4 driver is expressed in cardiac cells and pericardial cells but not in lymph gland cells. The expression profiles of HandD-gal4>GFP and NP1029-gal4>GFP in L1, L2 and L3 larval lymph glands are now provided in Figure 1—figure supplement 1 A-F’. Complementary information concerning these drivers is added in Materials and methods.

In all experiments, crosses and the subsequent raising of larvae until late L1/early L2 stage were performed at 22°C, before shifting larvae to 29°C until their dissection at the L3 stage. This information is now given both in the Materials and methods and the Results sections. Furthermore, according to the reviewer’s advice we performed temporal analysis of the lymph gland phenotype using tub-gal80ts (Figure 2—figure supplement 1O-Q) and looked at GCaMP3-expression at the L2 stage (Figure 5—figure supplement 1A-C). In conclusion, bnl/btl-FGF signaling is not involved in progenitor development till L2 but is required in third instar larvae to control the balance between MZ progenitor maintenance and hemocyte differentiation. This is now clearly stated in the revised manuscript.

b) Temporal role of Bnl in progenitor cells: The two Bnl sources, cardiac and blood cells, loss of Bnl in either gives the same phenotype. It is important to address the temporal requirement of Branchless and the dependence or independence of one source over the other to highlight the importance of the cardiac niche in establishing progenitor homeostasis. In its current form, the manuscript fails to highlight the importance of this niche. A comparative analysis of Bnl-GFP expression during lymph gland development (from early to late3rd instar) should be undertaken to reveal its expression profile within the cardiac cells and blood progenitor cells. Secondly, changes in Bnl-GFP pattern upon expressing BnlRNAi in cardiac cells or using Dome-gal4 will hopefully address the important contribution of cardiac niche in regulating progenitor Bnl levels. Finally, with regards to the role of Bnl, the other conceptual concern that is raised is its requirement either as a progenitor development signal or that it is required post progenitor development only as a maintenance cue. Again, using gal80ts as mentioned in the previous comment, should help clarify this aspect.

A detailed expression profile of Bnl is now provided. Since Bnl is diffusible, we performed ISH with simultaneous immunostainings with markers for different lymph gland cell types. These new data are provided in Figure 2a-a” and in Figure 2—figure supplement 1A-C”. We also looked at endogenous Bnl protein using the bnl:GFPendo knock-in alleles described by Du et al., 2018 (please see Figure 2S-T’).

When bnl was knocked-down in the cardiac tube (handD>bnl RNAi, bnl:GFPendo ) lower Bnl:GFPendo levels were recorded in MZ progenitors (Figure 2S-U). This result illustrates that Bnl secreted by cardiac cells contributes to Bnl levels in progenitors. These novel data are provided in Figure 2 and discussed in the text.

Concerning the temporal requirement of Bnl/Btll-FGF signaling, please see above our answer to point a.

c) Calcium homeostasis and FGF signaling: Although the authors show a down-regulation of GCAMP3 expression in lymph gland progenitor cells upon loss of FGF signaling, the analysis has been mostly done in the 3rd instar lymph gland when most of the tissue is differentiated. Hence it is hard to predict if the down-regulation is indeed because of loss of FGF signaling or is a consequence of progenitor differentiation and loss of these cells that maintain elevated GCAMP3 expression. Analysis of GCAMP3 levels in 2nd instar lymph glands prior to the onset of differentiation will help resolve this matter. Secondly, over-expression of PLC-γ in BtlRNAi condition may be important to prove the connection between Btl and activation of its downstream cascade linking to Calcium homeostasis more affirmatively.

As proposed by the reviewer, we looked at lymph gland GCaMP3 levels in second instar larvae. No difference compared to controls was observed (Figure 5—figure supplement 1 A-C), indicating that there is no difference in MZ progenitors till L2 stage.

Thanks to the reviewer’s remark about the connection between Btl and PLC-, we examined the relationship between sl and the Btl-FGF pathway and performed epistasis experiments (Figure 5J, K-N). Our data establish that sl acts downstream of the Bnl/Btl-FGF pathway.

Altogether, these data support the conclusion that in MZ progenitors, Bnl/Btl-FGF signaling leads to the activation of PLCg, which controls Ca2+ levels and in turn hemocyte differentiation. These new data are provided in (Figure 5J, K-N), discussed in the text and indicated on the model in Figure 6.

d) Cardiac niche to maintain progenitor heterogeneity: The authors talk about heterogeneity in progenitors and the importance of the cardiac niche in this "Furthermore, the MZ progenitor population is heterogeneous and a subset of progenitors, called "core progenitors", which express […] these data led us to ask whether signals derived from cardiac cells were involved in the control of lymph gland homeostasis, i.e.: the balance between progenitors and differentiated blood cells, independently from the PSC". This idea doesn't seem to crystallize in the course of the findings made. The readouts for the assays done are looking at crystal cell differentiation and progenitor maintenance status to decipher FGF signaling in progenitors with ligand contribution from cardiac cells. There needs to be some way to reconcile this either experimentally (differential effects on Tep4 and Dome expression under some genetic manipulations already shown in the manuscript) or textually in the Discussion.

Our apologies for not being clear on this point. It has been previously established that lymph gland MZ progenitors are labelled by either dome-MESO-LacZ (Krzemien et al., 2007) or domeMESO-RFP (Louradour et al., 2017). Col and tep4 are expressed in a subset of MZ progenitors called “the core progenitors”, which also express domeMESO-LacZ or domeMESO-RFP (Oyallon et al., 2016). In this study we carefully distinguished both progenitor types using the corresponding markers (DomeMESO-RFP and tep4 and/or Col). Knocking down bnl in cardiac cells or btl in MZ progenitors leads to a decrease in both DomeMESO-RFP and Col/tep4 expression, establishing that both progenitor types are affected. We modified the text in order to clarify this point.

Reviewer #3:

In this manuscript, Destalminil-Letourneau et al. describe a novel mechanism of progenitor maintenance in the hematopoietic organ, the lymph gland. They provide evidence for the presence of a vascular niche via cardiac cells, that regulates blood progenitor maintenance in the lymph gland. This manuscript provides interesting and novel insights by showing that the existence of a vascular niche as a conserved mechanism of blood stem cell maintenance as it also exists in flies. The authors provide mechanistic insight by providing data that supports the assertion that vascular niche-lymph gland mediated FGF pathway (via Bnl-Btl) signalling positively regulates hematopoietic progenitor maintenance by controlling intracellular calcium levels in the medullary zone of the lymph gland. The manuscript reports some exciting findings but could benefit from a few improvements requiring further experimentation, validation, and analysis. Major comments are as follows:

1) In Figure 2A-A' the authors that the Bnl signal is present throughout the primary lymph gland lobe. This raises the possibility that cells in Posterior Signaling Centre or the Cortical Zone can also be a source of Bnl for the prohemocytes via a type of reciprocal signalling. One way to look at the possibility that the PSC cells or CZ cells are a potential source of Bnl is to include high resolution images where Bnl is detected (either via antibody or in situ) and simultaneously label PSC cells (with collier or Antp-Gal4 driven mCD8GFP) or CZ cells (for example with Hmldelta-Gal4 driven mCD8GFP) with an in-situ against Bnl or with Bnl antibody – to rule out the possibility.

A detailed expression profile of Bnl is provided. Since Bnl is diffusible, we performed ISH with simultaneous immunostainings with specific markers for the different lymph gland cell types (PSC, MZ, crystal cell and differentiating hemocytes (Hml>GFP)). These new data are provided in Figure 2-A-A” and in Figure 2—figure supplement 1A-C”. We also looked at endogenous Bnl expression using the bnl:GFPendo knock-in alleles described by Du et al., 2018 (please see Figure 2S-T’). For Btl, we used the btl:mcherryendo knock-in allele generated by Du et al., 2018 and performed co-staining with markers for different lymph gland cell types (PSC, MZ, crystal cell and differentiating hemocytes (Hml>GFP)). These new data are provided in Figure 3A-A” and Figure 3—figure supplement 1 A-C”.

2) The author's arguments about tissue specific requirement of Bnl would be strengthened by testing whether the increased differentiation in bnlP2 mutants can be rescued by restoring bnl levels by transgene-mediated expression in the cardiac tube and the MZ. For example it could be determined whether the constitutive activation of Btl in prohemocytes in the genetic background of Btl mutants used (btldev1/+) rescues the crystal cell and plasmatocyte differentiation. Additionally, the authors could check if expressing Bnl in the MZ using dome or tep4-Gal4 can rescue the BnlRNAi phenotype of increased prohemocyte differentiation.

We expressed simultaneously bnl-RNAi and bnl in MZ progenitors (dome-gal4>bnl-RNAi>bnl) and observed a rescue in crystal cell numbers compared to the reduction of bnl alone (please see the quantification below). This result confirms that bnl-RNAi is specific to bnl, but this result does not say much about the relative contribution of bnl provided by MZ progenitors and bnl secreted by cardiac cells since we perform an overexpression of bnl in the MZ. As we already include many supplementary figures and as this information is not essential we did not integrate this result in the manuscript. However, what is more informative relative to the bnl source (CT versus MZ) are the novel data given in Figure 2S-U. We looked at endogenous Bnl using the bnl:GFPendo knock-in allele (Du et al., 2018). When bnl was knocked-down in the cardiac tube (handD>bnl RNAi, bnl:GFPendo ), lower Bnl:GFPendo levels were recorded in MZ progenitors (Figure 2S-U). This result illustrates that Bnl secreted by cardiac cells contributes to Bnl levels in progenitors. These novel data are provided in Figure 2 and discussed in the text.

3) Figure 3A also shows strong expression of Btl receptor in the cells towards the cortical zone (periphery of the LG). This raises some intriguing mechanistic questions. First, the manuscript would greatly benefit with a better, systematic, analysis of Bnl and Btl expression both in each of the zonal compartments with appropriate markers (PSC, MZ and CZ) – with high magnification/high resolution images. Second, One can envision a scenario where are the cells in the CZ themselves act as a source of Bnl that binds to Btl in neighboring (MZ) cells to regulate differentiation. Currently the authors have not ruled out this alternative model. If the Bnl is indeed secreted by differentiated cells then there could be 2 modes of signalling – either cell autonomous/paracrine mode or reciprocal signalling to maintain the MZ. It would be helpful if further clarification is provided of the role played by Bnl/Btl-FGF signalling in the CZ. It is possible is that the Bnl secreted by the cardiac cells is transcytosed/transported to the differentiated cells where it regulates differentiation cell autonomously. The authors could test this by perturbing Bnl and Btl in the CZ (using Hmldelta-Gal4 or eater-Gal4 for plasmatocytes and lz-Gal4 for crystal cells) and asking whether this affects crystal cell or plasmatocyte differentiation cell autonomously.

We have now established a detailed expression pattern of bnl and Btl using markers for the different LG compartments (please see above our response to point 1). Following the reviewer’s recommendation, we have studied the consequence of reducing bnl levels in crystal cells and in differentiating hemocytes with the specific Lz-Gal4 and HmlGal4 drivers, respectively. A decrease in mature crystal cells (labelled by Hnt) and mature plasmatocytes (labelled by P1) is observed when Lz-gal4 and Hml-gal4, respectively, are used to express bnl-RNAi (please see below the data). This result reveals a cell-autonomous function of bnl in these cells. These phenotypes are opposite to those due to bnl or btl reduction in cardiac cells and MZ progenitors, respectively. Multiple and distinct functions depend on the cellular context. Since this manuscript is centered on the communication between the cardiac tube and MZ progenitors, for the sake of clarity and to avoid a dilution of the main message, we choose to not present these data in the revised manuscript.

4) As noted by the authors it was previously shown that another FGF pathway in Drosophila, mediated by Heartless, is required for regulating hematopoiesis in the lymph gland (Dragojlovic-Munther et al., 2013). In this previous report overexpressing Btl in the progenitors had no effect (Figure S6B and K of that paper), this merits some discussion ad explanation from the authors.

In the paper Dragojlovic-Munther et al., 2013, no data were reported relative to the overexpression of the receptor breathless (Btl) in lymph gland progenitors. For Figure S5B of this paper, the authors performed the overexpression of the ligand bnl in progenitors. They measured the % of pxn labelled cells (pxn is a Cortical Zone marker) relative to all lymph gland cells and did not observe a significant difference compared to control. Unfortunately they did not analyze crystal cell or MZ progenitor indexes which might have been more sensitive reporters of the phenotype than pxn expression.

In our study, we used 2 independent UAS-Bnl transgenic lines (UASbnl described in Jarecki et al., 1999 and UAS-Bnl::GFP from Lin and Affolter, 2009) and the UAS-BtlCA (Pares and Ricardo, 2016). All 3 gave similar results (see Figure 2I; Figure 3G and Figure 2—figure supplement 2I).

Furthermore, since both Btl and Heartless signalling work through the transcription effector gene pointed we would expect cross talk between the two FGF pathways. This raises an important question of how the two pathways coordinated in the lymph gland. Some genetic interaction analysis between the components of the two arms of FGF signalling will be particularly useful. Specifically, determining if one of the FGF pathways is epistatic to the other, as well as establishing whether indeed both pathways converge on Pointed.

Thank you to the reviewer for her/his suggestion. To investigate the hierarchy between Htl/FGF and Btl/FGF pathways in MZ progenitors we performed epistasis experiments. Our results suggest that there is no hierarchy between these pathways and that both are required simultaneously in MZ progenitors to control lymph gland hematopoiesis. These new data are given in (Figure 3N-P) and are discussed in the text.

5) The Rab11 data in Figure 4B-B' is not as strong as it could be. First, the addition of a cell membrane marker and images of a better and higher resolution would allow greater clarity. Also, since Rab11 marks the recycling endosome, it would be more appropriate to look at other endocytic markers like Rab5 (early endosome marker) and/or Rab7 (late endosome) in addition to Rab11. Also, the co-expression data in the above cases with the Rabs is currently qualitative. It is appropriate in this sort of instance to include quantitative measurement of co-localization. The authors should consider including a staining showing endocytic localization (marked with the Rabs) of the Btl receptor (by in-situ or by antibody) in the prohemocytes. Also in Figure 4B-B' shows only two cells which have Rab11-bnl co-expression/co-localization whereas the field of interest seems to have many more cells in the area. Why don't all the prohemocytes show this co-expression? Finally, also in Figure 4 Bnl diffusion data in a-a' does not look convincing without a cellular marker to identify the hemocyte populations in the LG. Such a marker should be added.

In agreement with the reviewer’s comment we better defined bnl::GFP propagation from cardiac cells to MZ progenitors. We used the ubiRab11-cherryFP reporter and performed Rab7 immunostaining, which label recycling vesicles and late endosomes, respectively. Unfortunately we cannot look at Rab5 vesicles since the Rab5 reporter is tagged by GFP and therefore colocalisation with Bnl::GFP cannot be done. We also performed Col immunostaining to visualize core progenitors. These additional data are provided in Figure 4 A-A”. In addition, we looked at Bnl::GFP co-localisation with the Btl receptor (Figure 4 B-B”) and finally at Bnl::GFP diffusion from the CT to the CZ when the secretion of cardiac cells was impaired by sar1-RNAi expression. These data are provided in Figure 4 E-G. Altogether, they strongly support the proposition that Bnl::GFP secreted by cardiac cells is internalized in MZ progenitors likely through receptor-mediated endocytosis. This is discussed in the revised version of the manuscript.

https://doi.org/10.7554/eLife.64672.sa2

Article and author information

Author details

  1. Manon Destalminil-Letourneau

    Centre de Biologie du Développement (CBD), Centre de Biologie Intégrative (CBI), Université de Toulouse, CNRS, UPS, Toulouse, France
    Present address
    CellProthera, Paris, France
    Contribution
    Conceptualization, Formal analysis, Investigation, Methodology, Writing - original draft
    Competing interests
    No competing interests declared

  2. Ismaël Morin-Poulard

    Centre de Biologie du Développement (CBD), Centre de Biologie Intégrative (CBI), Université de Toulouse, CNRS, UPS, Toulouse, France
    Contribution
    Conceptualization, Formal analysis, Investigation, Methodology, Writing - review and editing
    Competing interests
    No competing interests declared

  3. Yushun Tian

    Centre de Biologie du Développement (CBD), Centre de Biologie Intégrative (CBI), Université de Toulouse, CNRS, UPS, Toulouse, France
    Contribution
    Formal analysis
    Competing interests
    No competing interests declared

  4. Nathalie Vanzo

    Centre de Biologie du Développement (CBD), Centre de Biologie Intégrative (CBI), Université de Toulouse, CNRS, UPS, Toulouse, France
    Contribution
    Formal analysis
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-6659-0299

  5. Michele Crozatier

    Centre de Biologie du Développement (CBD), Centre de Biologie Intégrative (CBI), Université de Toulouse, CNRS, UPS, Toulouse, France
    Contribution
    Conceptualization, Data curation, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Writing - original draft, Project administration, Writing - review and editing
    For correspondence
    michele.crozatier-borde@univ-tlse3.fr
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-9911-462X

Funding

Ministère de l'Education Nationale, de l'Enseignement Superieur et de la Recherche ('ANR programme blanc')

  • Michele Crozatier

Fondation ARC pour la Recherche sur le Cancer (ARC)

  • Manon Destalminil-Letourneau

Fondation pour la Recherche Médicale (FRM)

  • Michele Crozatier

Ligue Contre le Cancer

  • Michele Crozatier

China Scholarship Council

  • Yushun Tian

Société francaise d'hématologie

  • Manon Destalminil-Letourneau

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

We thank M Affolter, Y Bellaiche, J Casanova, J Colombani, L Du, M Freeman, M Grammont, M A Krasnow, A Paululat, L Perrin, S Ricardo, S Roy, T Tanaka, P Thérond, Bloomington and Vienna Stock Center and the TRiP at Harvard Medical School for fly strains; I Ando, A Moore and T Trenczek for antibodies; L Bataillé, A Davy, G Lebreton, M Meister, C Monod, B Monnier and A Vincent, for critical reading of the manuscript. We are grateful to B Ronsin and S Bosch for assistance with confocal microscopy (Plateforme TRI); J Favier, V Nicolas and A Destenable for fly culture. Research in the authors’ laboratory is supported by the CNRS, University Toulouse III, Ministère de la Recherche (ANR « programme blanc »), ARC (Association pour la Recherche sur le Cancer), La Ligue contre le Cancer 31, La Société Française d’Hématologie (SFH), FRM (Fondation pour la Recherche Médicale) and the China Scholarship Council.

Senior Editor

  1. Utpal Banerjee, University of California, Los Angeles, United States

Reviewing Editor

  1. Bruno Lemaître, École Polytechnique Fédérale de Lausanne, Switzerland

Reviewer

  1. Guy Tanentzapf, University of British Columbia, Canada

Publication history

  1. Received: November 6, 2020
  2. Accepted: December 16, 2020
  3. Version of Record published: January 4, 2021 (version 1)

Copyright

© 2021, Destalminil-Letourneau et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 1,193
    Page views
  • 150
    Downloads
  • 15
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Manon Destalminil-Letourneau
  2. Ismaël Morin-Poulard
  3. Yushun Tian
  4. Nathalie Vanzo
  5. Michele Crozatier
(2021)
The vascular niche controls Drosophila hematopoiesis via fibroblast growth factor signaling
eLife 10:e64672.
https://doi.org/10.7554/eLife.64672

Categories and tags

Research organism

Further reading

    1. Developmental Biology

    MMP14 cleaves PTH1R in the chondrocyte-derived osteoblast lineage, curbing signaling intensity for proper bone anabolism

    Tsz Long Chu, Peikai Chen ... Kathryn Song Eng Cheah
    Research Article Updated

    Bone homeostasis is regulated by hormones such as parathyroid hormone (PTH). While PTH can stimulate osteo-progenitor expansion and bone synthesis, how the PTH-signaling intensity in progenitors is controlled is unclear. Endochondral bone osteoblasts arise from perichondrium-derived osteoprogenitors and hypertrophic chondrocytes (HC). We found, via single-cell transcriptomics, that HC-descendent cells activate membrane-type 1 metalloproteinase 14 (MMP14) and the PTH pathway as they transition to osteoblasts in neonatal and adult mice. Unlike Mmp14 global knockouts, postnatal day 10 (p10) HC lineage-specific Mmp14 null mutants (Mmp14ΔHC) produce more bone. Mechanistically, MMP14 cleaves the extracellular domain of PTH1R, dampening PTH signaling, and consistent with the implied regulatory role, in Mmp14ΔHC mutants, PTH signaling is enhanced. We found that HC-derived osteoblasts contribute ~50% of osteogenesis promoted by treatment with PTH 1–34, and this response was amplified in Mmp14ΔHC. MMP14 control of PTH signaling likely applies also to both HC- and non-HC-derived osteoblasts because their transcriptomes are highly similar. Our study identifies a novel paradigm of MMP14 activity-mediated modulation of PTH signaling in the osteoblast lineage, contributing new insights into bone metabolism with therapeutic significance for bone-wasting diseases.

    1. Biochemistry and Chemical Biology
    2. Developmental Biology

    O-GlcNAc glycosylation orchestrates fate decision and niche function of bone marrow stromal progenitors

    Zengdi Zhang, Zan Huang ... Hai-Bin Ruan
    Research Article Updated

    In mammals, interactions between the bone marrow (BM) stroma and hematopoietic progenitors contribute to bone-BM homeostasis. Perinatal bone growth and ossification provide a microenvironment for the transition to definitive hematopoiesis; however, mechanisms and interactions orchestrating the development of skeletal and hematopoietic systems remain largely unknown. Here, we establish intracellular O-linked β-N-acetylglucosamine (O-GlcNAc) modification as a posttranslational switch that dictates the differentiation fate and niche function of early BM stromal cells (BMSCs). By modifying and activating RUNX2, O-GlcNAcylation promotes osteogenic differentiation of BMSCs and stromal IL-7 expression to support lymphopoiesis. In contrast, C/EBPβ-dependent marrow adipogenesis and expression of myelopoietic stem cell factor (SCF) is inhibited by O-GlcNAcylation. Ablating O-GlcNAc transferase (OGT) in BMSCs leads to impaired bone formation, increased marrow adiposity, as well as defective B-cell lymphopoiesis and myeloid overproduction in mice. Thus, the balance of osteogenic and adipogenic differentiation of BMSCs is determined by reciprocal O-GlcNAc regulation of transcription factors, which simultaneously shapes the hematopoietic niche.

    1. Developmental Biology

    The slingshot phosphatase 2 is required for acrosome biogenesis during spermatogenesis in mice

    Ke Xu, Xianwei Su ... Hongbin Liu
    Research Article

    The acrosome is a membranous organelle positioned in the anterior portion of the sperm head and is essential for male fertility. Acrosome biogenesis requires the dynamic cytoskeletal shuttling of vesicles towards nascent acrosome which is regulated by a series of accessory proteins. However, much remains unknown about the molecular basis underlying this process. Here, we generated Ssh2 knock-out (KO) mice and HA-tagged Ssh2 knock-in (KI) mice to define the functions of Slingshot phosphatase 2 (SSH2) in spermatogenesis and demonstrated that as a regulator of actin remodeling, SSH2 is essential for acrosome biogenesis and male fertility. In Ssh2 KO males, spermatogenesis was arrested at the early spermatid stage with increased apoptotic index and the impaired acrosome biogenesis was characterized by defective transport/fusion of proacrosomal vesicles. Moreover, disorganized F-actin structures accompanied by excessive phosphorylation of COFILIN were observed in the testes of Ssh2 KO mice. Collectively, our data reveal a modulatory role for SSH2 in acrosome biogenesis through COFILIN-mediated actin remodeling and the indispensability of this phosphatase in male fertility in mice.