1. Developmental Biology
Download icon

Fatty acid β-oxidation is required for the differentiation of larval hematopoietic progenitors in Drosophila

  1. Satish Kumar Tiwari
  2. Ashish Ganeshlalji Toshniwal
  3. Sudip Mandal
  4. Lolitika Mandal  Is a corresponding author
  1. Developmental Genetics Laboratory, Department of Biological Sciences, Indian Institute of Science Education and Research (IISER) Mohali, India
  2. Molecular Cell and Developmental Biology Laboratory, Department of Biological Sciences, Indian Institute of Science Education and Research (IISER) Mohali, India
Research Article
  • Cited 0
  • Views 2,755
  • Annotations
Cite this article as: eLife 2020;9:e53247 doi: 10.7554/eLife.53247

Abstract

Cell-intrinsic and extrinsic signals regulate the state and fate of stem and progenitor cells. Recent advances in metabolomics illustrate that various metabolic pathways are also important in regulating stem cell fate. However, our understanding of the metabolic control of the state and fate of progenitor cells is in its infancy. Using Drosophila hematopoietic organ: lymph gland, we demonstrate that Fatty Acid Oxidation (FAO) is essential for the differentiation of blood cell progenitors. In the absence of FAO, the progenitors are unable to differentiate and exhibit altered histone acetylation. Interestingly, acetate supplementation rescues both histone acetylation and the differentiation defects. We further show that the CPT1/whd (withered), the rate-limiting enzyme of FAO, is transcriptionally regulated by Jun-Kinase (JNK), which has been previously implicated in progenitor differentiation. Our study thus reveals how the cellular signaling machinery integrates with the metabolic cue to facilitate the differentiation program.

eLife digest

Stem cells are special precursor cells, found in all animals from flies to humans, that can give rise to all the mature cell types in the body. Their job is to generate supplies of new cells wherever these are needed. This is important because it allows damaged or worn-out tissues to be repaired and replaced by fresh, healthy cells.

As part of this renewal process, stem cells generate pools of more specialized cells, called progenitor cells. These can be thought of as half-way to maturation and can only develop in a more restricted number of ways. For example, so-called myeloid progenitor cells from humans can only develop into a specific group of blood cell types, collectively termed the myeloid lineage.

Fruit flies, like many other animals, also have several different types of blood cells. The fly’s repertoire of blood cells is very similar to the human myeloid lineage, and these cells also develop from the fly equivalent of myeloid progenitor cells. These progenitors are found in a specialized organ in fruit fly larvae called the lymph gland, where the blood forms. These similarities between fruit flies and humans mean that flies are a good model to study how myeloid progenitor cells mature.

A lot is already known about the molecules that signal to progenitor cells how and when to mature. However, the role of metabolism – the chemical reactions that process nutrients and provide energy inside cells – is still poorly understood. Tiwari et al. set out to identify which metabolic reactions myeloid progenitor cells require and how these reactions might shape the progenitors’ development into mature blood cells.

The experiments in this study used fruit fly larvae that had been genetically altered so that they could no longer perform key chemical reactions needed for the breakdown of fats. In these mutant larvae, the progenitors within the lymph gland could not give rise to mature blood cells. This showed that myeloid progenitor cells need to be able to break down fats in order to develop properly.

These results highlight a previously unappreciated role for metabolism in controlling the development of progenitor cells. If this effect also occurs in humans, this knowledge could one day help medical researchers engineer replacement tissues in the lab, or even increase our own bodies’ ability to regenerate blood, and potentially other organs.

Introduction

Recent studies have highlighted how metabolism regulates the state and fate of stem cells (Ito and Suda, 2014; Shyh-Chang et al., 2013; Shyh-Chang and Ng, 2017). Besides catering to the bioenergetic demands of a cell, metabolic intermediates can also alter the fate of stem cells via epigenetic mechanisms like histone modifications (Atlasi and Stunnenberg, 2017; Saraiva et al., 2010). Studies on diverse stem cell scenarios, primarily in Hematopoietic Stem Cells (HSCs), have established that at various developmental stages, stem cells have different metabolic requirements (Kohli and Passegué, 2014; Suda et al., 2011). Nevertheless, the metabolic demand of progenitors, the immediate descendants of stem cells, is yet to be fully elucidated.

Studies to date have evidenced that glucose metabolism impacts the onset and magnitude of HSC induction (Harris et al., 2013; Shyh-Chang and Ng, 2017), as well as HSC specification (Oburoglu et al., 2014). Another major metabolic state that is active in stem and progenitor cells is fatty acid oxidation (FAO) (Ito et al., 2012; Knobloch et al., 2017; Lin et al., 2018; Wong et al., 2017; Ito et al., 2012). Enzymatic activities of the members of FAO lead to the shortening of fatty acids and the production of acetyl-CoA in mitochondria. The acetyl-CoA thus produced can not only generate NADH and FADH2 through TCA cycle but also can be utilized in acetylation of various proteins including histones (Fan et al., 2015; Houten and Wanders, 2010; McDonnell et al., 2016; Wong et al., 2017).

The primary goal of this study was to ascertain whether FAO regulates any aspect of the hemocyte progenitors in the Drosophila larval hematopoietic organ, lymph gland. The lymph gland is a multilobed structure consisting of a well-characterized anterior lobe (primary lobe) and uncharacterized posterior lobes (Figure 1A, Banerjee et al., 2019). The core of the primary lobe houses the progenitor populations and is referred to as the medullary zone (MZ), while the differentiated cells define the outer cortical zone (CZ, Figure 1A'). In between these two zones, lies a rim of differentiating progenitors or intermediate progenitors (IPs). The blood progenitors of late larval lymph gland are arrested in G2-M phase of cell cycle (Sharma et al., 2019), have high levels of ROS (Owusu-Ansah and Banerjee, 2009), lack differentiation markers, are multipotent (Jung et al., 2005) and are maintained by the hematopoietic niche/posterior signaling center, PSC (Krzemień et al., 2007; Lebestky et al., 2003; Mandal et al., 2007). The primary lobe has been extensively used to understand intercellular communication relevant to progenitor maintenance (Gao et al., 2013; Giordani et al., 2016; Gold and Brückner, 2014; Hao and Jin, 2017; Krzemień et al., 2007; Krzemien et al., 2010; Lebestky et al., 2003; Mandal et al., 2007; Mondal et al., 2011; Morin-Poulard et al., 2016; Sinenko et al., 2009; Small et al., 2014; Yu et al., 2018). Although these studies have contributed significantly toward our understanding of cellular signaling relevant for progenitor homeostasis, the role of cellular metabolism in regulating the state and fate of blood progenitors remains to be addressed.

Figure 1 with 2 supplements see all
FAO genes are expressed in hemocyte progenitors of lymph gland.

Age and genotype of the larvae are mentioned in respective panels. (A–A') Model of lymph gland of third early and third late instar stages depicting anterior primary lobes and posterior lobes. (A’). Primary lobe showing different subpopulations: Pvf2+ Dome- pre-progenitor, Dome+ progenitors and Dome+ Pxn+ HmlIntermediate progenitors (IPs) in early third and late third instar larval stages. Progenitors are present in the core of the primary lobe called the medullary zone (MZ), and differentiated cells (Plasmatocytes and crystal cells) are present in the outer zone called cortical zone (CZ). (B–B'') Expression of Hnf4-GAL4 > UAS-GFP in Pvf2+ pre-progenitors of the early third instar lymph gland. (C–C'') Expression of Hnf4-GAL4 > UAS-GFP in Dome+ progenitors and Dome+ HmlIntermediate progenitors (IPs) shown in dome-MESO-EBFP2/+; Hml-DsRed/+ genotype. (D). Quantitative analysis of B–C''- reveals that the Dome+ progenitors have higher levels of Hnf4 expression. p-Value for Hnf4-GAL4 > UAS-GFP expression in Dome+ progenitors is 9.55 × 10−9 compared to control Pvf2+ pre-progenitors. p-Value for Hnf4-GAL4 > UAS-GFP expression for Dome+ Hml+ IPs is 7.34 × 10−3 compared to control Pvf2+ pre-progenitors. (E–E'') Nile red staining in Pvf2+ pre-progenitors of early third instar stage lymph gland. (F–F'') Expression of Nile red in Dome+ progenitors and Dome+ HmlIntermediate progenitors (IPs) shown in dome-MESO-EBFP2/+; Hml-DsRed/+ genotype (Dome+: blue, Hml+: green). (G). Quantitative analysis of E–F'' shows higher levels of neutral lipids in the Dome+ progenitors. Compared to control Pvf2+ pre-progenitors, p-Values for nile red expression in Dome+ progenitors is 1.39 × 10−7 and Dome+ Hml+ IPs pre-progenitors is 9.11 × 10−3. Five optical sections of 1 µm thickness from the middle of the Z-stack were merged into a single section. (H) Schematic representation of FAO and the constituent enzymes. (I) Transcripts of β-oxidation enzymes, whd, Mcad, Mtpα, scully, Mtpβ, and yip2 (Refer to H) can be detected in the third late instar lymph gland. eL3 and lL3 refer to the early and late instar lymph glands. Individual dots represent biological replicates. Values are mean ± SD, asterisks mark statistically significant differences (*p<0.05; **p<0.01; ***p<0.001, Student’s t-test). Scale bar: 20 µm.

Here, we show that the G2-M arrested hemocyte progenitors of the Drosophila larval lymph gland rely on FAO for their differentiation. While the loss of FAO prevents their differentiation, upregulation of FAO in hemocyte progenitors by either genetic or pharmacological means leads to precocious differentiation. More importantly, acetate supplementation restores the histone acetylation and differentiation defects of the progenitor cells observed upon loss of FAO. Our genetic and molecular analyses reveal that FAO acts downstream to the Reactive Oxygen Species (ROS) and c-Jun N-terminal Kinase (JNK) axis, which is essential for triggering the differentiation of these progenitors (Owusu-Ansah and Banerjee, 2009). In this study, we, therefore, provide the unknown link that connects cellular signaling and metabolic circuitry essential for differentiation of the blood progenitors.

Results

Genes involved in FAO pathway are expressed in the hemocyte progenitors of late third instar lymph gland

Drosophila hemocyte progenitors in the lymph gland proliferate in the early larval stages (Jung et al., 2005; Mondal et al., 2011). Eventually, they undergo a G2-M arrest in late third instar (Sharma et al., 2019). Studies have identified that lymph gland hemocyte progenitors of the primary lobe can be grouped into three subpopulations: Dome- pre-progenitors, Dome+ progenitors, and Dome+ Pxn+ HmlIntermediate progenitors (Banerjee et al., 2019). These progenitor subpopulations will be henceforth referred to as pre-progenitors, progenitors, and IPs, respectively. The pre-progenitors can also be visualized by Pvf2 expression in first, second, and early third instar larval lymph gland (Ferguson and Martinez-Agosto, 2017, Figure 1—figure supplement 1A–B). Since in the late third instar lymph gland, only progenitors and IPs are present (Figure 1A'), we analyzed the early third instar lymph gland to characterize the pre-progenitors.

To ascertain the involvement of FAO, if any, in hemocyte progenitors of larval lymph gland, we monitored the expression of Hepatocyte Nuclear Factor 4 (Hnf4) (Palanker et al., 2009), an essential gene of larval FAO and lipid mobilization. As evident from Figure 1B–D, G2–M arrested progenitors (visualized by dome-MESO-EBFP2+) express high levels of Hnf4 > GFP compared to pre-progenitors (visualized by Pvf2 expression) and IPs (Dome+ Hml+). Interestingly, neutral lipids (visualized by Nile red staining: Figure 1E–G) are conspicuous in late third instar blood progenitors (dome-MESO-EBFP2+), compared to the pre-progenitors (Figure 1E–E'' and G) and IPs (Figure 1F–F'' and G ).The lipid enrichment in the late progenitors is further evident upon LipidTOX (validated marker for neutral lipids) labeling (Figure 1—figure supplement 2A–C). Based on the presence of relatively high levels of lipid droplets in a non-lipid storage tissue and the expression of Hnf4 in the progenitor cells, we speculated a developmental role of FAO in these cells. This prompted us to check for the expression of other genes involved in FAO (Palanker et al., 2009, Figure 1H). Figure 1I shows the expression of the rate-limiting enzyme of FΑO, withered (whd, Drosophila homolog of CPT1: Carnitine palmitoyltransferase 1) along with Mcad (medium-chain acyl-CoA dehydrogenase), Mtpα (mitochondrial trifunctional protein α subunit: Long-chain-3-hydroxyacyl-CoA dehydrogenase), scully (3-hydroxyacyl-CoA dehydrogenase), Mtpβ (mitochondrial trifunctional protein β subunit: Long-chain-3-hydroxyacyl-CoA dehydrogenase) and yip2 (yippee interacting protein 2: acetyl-CoA acyltransferase) in the late lymph gland. Elevated levels of expression of acyl-Coenzyme A dehydrogenase (CG3902) is also seen in the progenitors as compared to the pre-progenitors and IPs (Figure 1—figure supplement 2D–F).

Major aspects of FAO takes place in the mitochondria where fat moiety is broken down to generate acetyl-CoA, NADH, and FADH2 (Bartlett and Eaton, 2004), we, therefore, looked at the status of mitochondria in the progenitors as well as the differentiated hemocytes. The presence of an abundant reticular network of mitochondria is evident in the progenitors (dome-GAL4 >UAS-mito-HA-GFP, Figure 1—figure supplement 1C–C', and Video 1). However, for reasons unknown to us HmlΔ-GAL4 is unable to drive UAS-mito-HA-GFP, therefore, we used streptavidin labeling to visualize the mitochondrial status in the differentiated hemocytes. Interestingly, differentiated hemocytes (HmlΔ-GAL4 > UAS-GFP) show less reticular mitochondria (labeled by Strepatvidin-Cy3, Chowdhary et al., 2017; Hollinshead et al., 1997, Figure 1—figure supplement 1D–D') in comparison to the progenitors (Hml>GFP negative) thereby indicating a preference for FAO in the hemocyte progenitors.

Video 1
Mitochondrial distribution in the progenitors (red, Dome+) visualized by UAS-mito-HA-GFP.

Put together; the above results implicate FAO as the metabolic state of the Dome+ cells of the primary lobe of the lymph gland. These observations, in turn, encouraged us to investigate the importance of FAO in maintaining the state and fate of these progenitors during development.

Loss of FAO affects hemocyte progenitor differentiation

Drosophila ortholog of CPT1, withered (whd, Figure 2A), is a rate-limiting enzyme for FAO (Strub et al., 2008). The loss of function of CPT1/whd, therefore, blocks mitochondrial FAO (Schreurs et al., 2010). To investigate the role of FAO in late hemocyte progenitors, we employed a null allele of withered, whd1 (Strub et al., 2008). The primary lobes of whd1 homozygous lymph gland have abundant Dome+ progenitors but drastically reduced number of differentiated hemocytes (P1: plasmatocytes Figure 2B–C and D; proPO: Crystal cells, Figure 2E–F and G), and Intermediate progenitors (Dome+ Pxn+, Figure 2H–I'' and Figure 2—figure supplement 1A) compared to control.

Figure 2 with 2 supplements see all
Loss of fatty acid β-oxidation affected differentiation of hemocyte progenitors of the lymph gland.

(A) Schematic representation of fatty acid β-oxidation within the mitochondria of a cell. (B–D) Compared to control (B) decrease in differentiation (red, reported by P1 immunostaining) and increment in progenitor number (dome > GFP) is observed in the lymph gland of a homozygous null allele of whd (C). (D) Quantitative analysis of B–C reveals a significant increment in the number of Dome+ progenitors. p-Value for dome-GAL4, UAS-GFP; whd1/whd1=2.67×10−10 compared to control. (E–G) Compared to control (E) decrease in crystal cell number (red, reported by proPO immunostaining) and increment in the progenitor cell population (dome > GFP) is observed in the lymph gland of the homozygous null allele of whd (F). (G). Quantitative analysis of results from E–F shows a significant drop in the number of crystal cells. p-Value for dome-GAL4, UAS-GFP; whd1/whd1=4.38×10−7 compared to control. (H–I'') The hemocyte progenitor subpopulation dynamics (red, reported by Pxn immunostaining and green marking dome > GFP) of Dome+ progenitors and Dome+ Pxn+ (IPs) in the late third instar lymph gland of control (H–H'') and homozygous null allele of whd (I–I''). (J–S) Spatio-temporal analysis of hemocyte progenitor subpopulations of Dome- pre-progenitors, Dome+ progenitors, and Dome+ Pxn+ (IPs) (red, reported by Pxn immunostaining and green marking dome > GFP) observed in the lymph gland of control (J–N) and homozygous null allele of whd (O–S). Insets in K, L, and M show pre-progenitors, progenitors and intermediate progenitors respectively in control and inset in Q shows abundant progenitors in the homozygous null allele of whd. (T–X) Compared to control (T) decrease in differentiation (red, reported by P1 immunostaining) and increase in progenitor number (dome > GFP) is observed in lymph gland upon progenitor specific RNAi based down-regulation of whd (U) CRISPR-Cas9 based knock-out of whd (V) and miRNA based knockdown of Hnf4 (W). (X) Quantitative analysis of the results from T–W, illustrating the significant increase in Dome+ progenitors upon targeted loss of FAO. p-Value for dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-whd RNAi = 2.84×10−15 compared to control. p-Value for dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-dCas9; U-6: sgRNA-whd = 3.84×10−19. p-Value for dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-Hnf4.miRNA =6.04×10−14. Individual dots represent biological replicates. Values are mean ± SD, asterisks mark statistically significant differences (*p<0.05; **p<0.01; ***p<0.001, Student’s t-test). Scale bar: 20 µm.

Detailed temporal analysis of the dynamics of progenitor subpopulation during normal development, as well as upon loss of whd1 (Figure 2J–S), was next carried out. In sync with an earlier report (Ferguson and Martinez-Agosto, 2017), our analysis reveals that Dome- pre-progenitors (Figure 2K) are present in the developing lymph gland until the early third instar (Figure 2L and Figure 2—figure supplement 1B). Beyond this timeline, the subsets that populates the lymph gland are Dome+ progenitors (Figure 2M–N and Figure 2—figure supplement 1B), and Dome+ Pxn+ IP cells (Figure 2M–N and Figure 2—figure supplement 1B). Interestingly, in whd1 mutant lymph glands, while the pre-progenitors are present till the early third instar stage (Figure 2P and Figure 2—figure supplement 1C), there is an abundance of progenitors (Figure 2Q–S) with a small number of IP cells (Figure 2R–S). Quantification of the above results reflects that in whd1 mutant the progenitors are rather stalled instead of undergoing a natural transition to IP cells with time (Figure 2—figure supplement 1C). During normal development, the first sign of differentiation (as evidenced by Pxn expression) occurs around 36 hr AEH, and by 96 hr AEH, there is a prominent cortical zone defined by the differentiating cells.

In contrast, the cortical zone is drastically reduced due to lack of differentiation in the whd1 mutant lymph gland. The timed analysis also revealed that the defect seen in differentiation in this mutant has an early onset (36 hr AEH, Compare Figure 2O with Figure 2J). These observations implicate that the lack of FAO dampens the differentiation process of Dome+ progenitors.

Since the differentiation of progenitors is affected, we next performed an RNAi-mediated downregulation of whd by the TARGET system (McGuire et al., 2004; Figure 2—figure supplement 1D).

Progenitor-specific downregulation (dome-GAL4, UAS-GFP; tubGAL80ts20; UAS-whd RNAi) results in a halt in differentiation, as evidenced by an increase in the area of dome > GFP and a concomitant decline in adjoining CZ (visualized by differentiated plasmatocyte (Nimrod: P1, Figure 2U) compared to control (Figure 2T). Upon activation of whd RNAi from a different source (VDRC) by another independent progenitor specific driver TepIV-GAL4 (Figure 2—figure supplement 1E–G), a similar result is obtained. Additionally, progenitor-specific knockout of whd by CRISPR/Cas9 system (Hsu et al., 2014) supports the above results (Figure 2V and X).

Likewise, knockdown of the key player in lipid metabolism Hnf4 results in a decline in the differentiation of the progenitors, endorsing the role of FAO in progenitor differentiation (Figure 2W and X, Figure 2—figure supplement 2). Lymph glands from a hetero-allelic combination of dHNF4 (dHNF4Δ17/dHNF4Δ33: null allele of Hnf4) (Palanker et al., 2009), exhibits an abundant progenitor pool (Shg: DE-Cadherin) coupled with the reduction in differentiated cells (Pxn, compare Figure 2—figure supplement 1I with 1H, and Figure 2—figure supplement 1M), further denoting that FAO disruption indeed leads to compromised differentiation of progenitors.

To further verify our observations, we next analyzed the homozygous mutant of two essential enzymes of β−oxidation: Mtpα (mitochondrial trifunctional protein α subunit: Long-chain-3-hydroxyacyl-CoA dehydrogenase), and Mtpβ (mitochondrial trifunctional protein β subunit: Long-chain-3-hydroxyacyl-CoA dehydrogenase) (Kishita et al., 2012). Primary lobes from homozygous Mtpα[KO] and Mtpβ[KO] loss of function has a large progenitor pool (Shg: DE-Cadherin) at the expense of differentiated cells (Pxn) (compare Figure 2—figure supplement 1J–K with H), a phenotype similar to the whd1 (Figure 2—figure supplement 1L and M). DE-cadherin enrichment in the FAO loss of function progenitors is also indicative of high-maintenance signals (Gao et al., 2014).

In addition to genetic knockdown and classical loss-of-function analyses, we performed pharmacological inhibition of FAO in Drosophila larvae (Figure 2—figure supplement 1N–Q). Larvae grown in food supplemented with FAO inhibitors, Etomoxir (Lopaschuk et al., 1988), and Mildronate also demonstrate a more than two-fold reduction in the differentiation of the progenitors of the primary lobe.

Collectively, our genetic and pharmacological studies illustrate the cell-autonomous role of FAO in the differentiation of blood progenitors of the lymph gland.

Loss of FAO causes an increase in progenitor proliferation

Previous study from our laboratory demonstrated that G2-M arrest is a hallmark of the otherwise proliferating progenitors prior to differentiation (Sharma et al., 2019). In contrast, we found that whd1 homozygous progenitors exhibit a stark increase in EdU incorporation implicating their highly proliferating status upon loss of FAO when compared to age-isogenised controls (Figure 3A–C). To have a further insight into the cell cycle status, we expressed Fly-FUCCI-fluorescent ubiquitination-based cell cycle indicator, specifically in the progenitors. This indicator employs two fluorescent probes: the first probe is an E2F moiety fused to GFP, which is degraded by Cdt2 during the S-phase. This construct allows the visualization of G1, G2, and M-phase cells by GFP expression. The second probe is a fusion of Red Fluorescent Protein (mRFP.nls) and CycB moiety. It is degraded by the anaphase-promoting complex/cyclosome (APC/C) during midmitosis, thereby reporting the cell in S, G2, and M-phase. Together this system allows the visualization of cells in G1, S, and G2/early mitosis by green, red, and yellow signals, respectively (Zielke et al., 2014).

Figure 3 with 1 supplement see all
Loss of fatty acid β-oxidation causes an increase in proliferation of hemocyte progenitors of the lymph gland primary lobe.

Genotype of the larvae are mentioned in respective panels. (A–C) The difference in proliferation status (reported by EdU incorporation) in the lymph gland of third late instar control larvae (A–A') compared to whd null mutant (B–B'). (C). Quantitative analysis of the results from A–B' illustrates a significant increase in proliferation in whd1 Dome+ progenitors. p-Value for dome-GAL4, UAS-GFP; whd1/whd1=1.71×10−6 compared to control. (D–F) Difference in cell cycle status (reported by Fly-FUCCI) in the lymph gland of third late instar control larvae (D) compared to the progenitor-specific RNAi-based down-regulation of whd (E). (D'–E'): Pie chart depicting the fraction of G1 (green), S(red), and G2/M (yellow) progenitors in J–K. (F) Quantitative analysis of the results from D–E. p-Value for red cells in dome-GAL4, UAS-Fly-FUCCI; UAS-whd RNAi = 1.41×10−11, p-Value for green cells in dome-GAL4, UAS-Fly-FUCCI; UAS-whd RNAi = 2×10−4, p-Value for yellow cells in dome-GAL4, UAS-Fly-FUCCI; UAS-whd RNAi = 1.5×10−5 compared to control. ns.=not significant, Individual dots represent biological replicates. Values are mean ± SD, asterisks mark statistically significant differences (*p<0.05; **p<0.01; ***p<0.001, Student’s t-test). Scale bar: 20 µm.

We found that, instead of being in a G2-M arrest, the hemocyte progenitors in whd mutant are actively proliferating, as evidenced by more cells in S-phase (red) compared to the control (Figure 3D–D' with E-E'). Quantitative analyses reveal more than three-fold increase in the number of cells in the S phase with a concomitant drop in G2-M arrested progenitors (compare Figure 3D' with E' and F).

Together these results assert that FAO disruption in proliferating progenitors doesn’t allow them to halt at G2-M and subsequently differentiate.

Failure in differentiation of hemocyte progenitors upon loss of FAO is not due to decline in ROS levels

The perturbation of differentiation prompted us to look at the status of both differentiation and maintenance factors, per se in these hemocyte progenitors. Although the progenitor pool in the larval lymph gland is heterogenous (Baldeosingh et al., 2018; Banerjee et al., 2019; Sharma et al., 2019), our timed analysis indicates that the majority of the progenitors populating the late third instar lymph gland expresses dome. Hedgehog signaling has been implicated in the maintenance of the dome expressing progenitors (Baldeosingh et al., 2018; Mandal et al., 2007; Sharma et al., 2019; Tokusumi et al., 2010). Hematopoietic niche/PSC releases Hh, which leads to the expression of the Hh signal transducer Cubitus interruptus (Ci155[Alexandre et al., 1996]) in the progenitors. The homozygous whd1 lymph gland progenitors express a higher level of Ci155 compared to control (compare Figure 3—figure supplement 1B-B'' with A-A'' and quantitated in Figure 3—figure supplement 1C) correlating with higher proliferation and less differentiation (Figure 2B–D). This observation, along with enrichment of DE-cadherin in FAO mutants (Figure 2—figure supplement 1H–M), endorses high-maintenance signal in the progenitors.

Reactive oxygen species (ROS) is the major signal attributed to the differentiation of the hemocyte progenitors of the lymph gland (Owusu-Ansah and Banerjee, 2009). High levels of developmentally generated ROS trigger Jun Kinase (JNK) signaling, which sets these progenitors toward the differentiation program (Owusu-Ansah and Banerjee, 2009). Since homozygous whd1 progenitors fail to differentiate, we rationalized that this might be due to the drop in the differentiation signal ROS.

To probe this possibility, we analyzed the levels of ROS in whd1 lymph glands by dihydroxy ethidium (DHE) staining. Quite strikingly, whd1 homozygous progenitors exhibited elevated levels of ROS (Compare Figure 3—figure supplement 1D–D'' with 3E-E'' and quantified in Figure 3—figure supplement 1F). A similar observation of increased ROS in CPT1 knockdown endothelial cells has been reported in another study (Kalucka et al., 2018). Therefore, we can infer that the halt in differentiation observed in the Ci155 enriched progenitors of homozygous whd1 is not an outcome of compromised ROS level.

Collectively, these results indicate that despite achieving a high ROS level than the control, the progenitors are unable to move into differentiation, indicating that FAO might act downstream to ROS.

Upregulation of FAO causes precocious G2-M arrest and differentiation of blood progenitors

Due to the central role of carnitine in fat metabolism, it is often used as a supplement for enhancing fat oxidation (Pekala et al., 2011; Wall et al., 2011). Several studies have concluded that FAO can be upregulated in the cells by L-carnitine supplementation (Pekala et al., 2011; Sahlin, 2011; Wall et al., 2011).

Our loss-of-function genetic analyses illustrated that FAO is essential for the differentiation of hemocyte progenitors. Whether it is sufficient for the progenitor differentiation was addressed by feeding L-carnitine supplemented food (100 mM concentration for 48 hr) to wandering third instar larvae. Compared to control (Figure 4A), the lymph gland from L-carnitine fed larvae (96 hr AEH) exhibits a drastic reduction in progenitor zone (visualized by dome > GFP) with a concomitant increase in differentiation (visualized by P1: Nimrod; Figure 4B–C). The increase in differentiation by L-carnitine supplementation is also apparent in whd1 heterozygous mutants. Feeding L-carnitine could rescue the differentitaion defects associated with whd1 heterozygous mutants (Compare Figure 4—figure supplement 1A–E).

Figure 4 with 1 supplement see all
FAO upregulation results in precocious differentiation and G2 arrest in hemocyte progenitors.

Age and genotype of the larvae are mentioned in respective panels. (A–C) Comparison of differentiation (marked by P1) levels in dome > GFP lymph gland of control (A) and L-carnitine supplemented (B) larvae. (C) Quantitative analysis of results from A–B showing increased differentiation upon L-carnitine supplementation. p-Value for dome-GAL4, UAS-GFP = 2.37×10−10 supplemented with L-carnitine compared to control. (D–F) Comparison of differentiation (marked by P1) levels in dome > GFP lymph gland of control (D) and CRISPR-Cas9 mediated whd overexpression (E) in larval hemocyte progenitors. (F) Quantitative analysis of result from D–E depicting a significant increase in differentiation upon overexpression of whd. p-Value for dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-whd-OE = 5.82×10−12 compared to control. (G–H) Comparison of differentiation (marked by P1) levels in dome > GFP lymph gland secondary lobes (marked by the white dotted boundary) of control (G) and CRISPR-Cas mediated whd overexpression (H) in larval hemocyte progenitors. (I–K) Proliferation status (marked by EdU) in third early instar hemocyte progenitors (dome > GFP) of control (I–I') and L-carnitine supplemented (J–J') larvae. (K) Quantitative analysis of results from I–J' reveals a decline in the number of proliferating Dome+ progenitors upon FAO overexpression. p-Value for dome-GAL4, UAS-GFP fed with L-carnitine = 2.87×10−5 compared to control. (L–N) The decline in proliferation status (marked by EdU) in third early instar hemocyte progenitors of CRISPR-Cas9 mediated whd overexpression (M–M') compared to (dome > GFP) of control (L–L'). (N) Quantitative analysis of result from L–M'. p-Value for dome-GAL4, UAS-GFP; tubGAL80ts20 > whd-OE=2.28×10−9 compared to control. (O–P) Alteration in cell cycle status (reported by Fly-FUCCI) in the lymph gland of third late instar larvae grown in L-carnitine supplemented food (P) compared to control (dome > UAS-FUCCI) (O). (O'–P'): Pie chart depicting the fraction of G1 (green), S (red), and G2/M (yellow) progenitors in (O-P). (Q) Quantitative analysis of the results from O–P, illustrating the increase in G2-M upon FAO overexpression. p-Value for red cells in L-carnitine supplemented dome-GAL4, UAS-FUCCI = 1.78×10−7, p-Value for green cells in L-carnitine supplemented dome-GAL4, UAS-FUCCI; UAS-whd RNAi = 2.16×10−1. p-Value for yellow cells in L-carnitine supplemented dome-GAL4, UAS- FUCCI; UAS-whd RNAi = 1.71×10−5 compared to control. Individual dots represent biological replicates. Values are mean ± SD, asterisks mark statistically significant differences (*p<0.05; **p<0.01; ***p<0.001, Student’s t-test). Scale bar: 20 µm.

As a genetic correlate, whd was overexpressed by a Cas9-based transcriptional activator (BDSC68139) (Ewen-Campen et al., 2017) in the hemocyte progenitors following the scheme in Figure 2—figure supplement 1D. Overexpression of whd indeed results in an increase in the differentiation of hemocyte progenitors (compare Figure 4D with Figure 4E–F). It is interesting to note that the otherwise undifferentiated reserve progenitors of the secondary lobes also differentiate upon whd overexpression (marked by a dotted white line in Figure 4G–H).

We next employed dual fly-FUCCI construct, and EdU labeling to assay the cell cycle cell status of the FAO upregulated progenitors. The early third instar lymph glands from L-carnitine fed larvae exhibit a radical decline in EdU incorporation compared to the proliferating progenitors of the control larvae of similar age (compare Figure 4I–I' with Figure 4J-J' and quantified in Figure 4K). As a genetic correlate, we overexpressed whd in the progenitors, which also led to a decline in EdU incorporation (Figure 4L–N). Our FUCCI analysis reveals an abundance of G2-M progenitors in the early third instar larvae reared in L-carnitine supplemented food compared to control samples. Thus, less EdU incorporation due to upregulated FAO resulted in an early onset of G2-M arrest in the progenitors (compare Figure 4O–O' with Figure 4P–P' and quantified in Figure 4Q).

Put together; our results reveal that upon FAO upregulation, the progenitor experiences a precocious G2-M halt in their cell cycle. During normal development, the late progenitors also undergo a G2-M halt before they differentiate. We, thus, infer that FAO is imperative for the differentiation of lymph gland progenitors.

FAO loss in hemocyte progenitors leads to sustained glycolysis

Next, we investigated whether a compromise in the intracellular energy source (ATP) is the reason behind the differentiation defect seen in the FAO mutants. Quite intriguingly, ATP levels of homozygous whd1 larvae are comparable to similarly aged control (Figure 5A). This unaltered ATP level is in sync with our observation of higher proliferation observed in homozygous whd1 hemocyte progenitors (Figure 3B–B'). Since higher proliferation is driven by elevated glycolysis in different scenarios (Lunt and Vander Heiden, 2011), we hypothesized that loss of FAO might push the progenitors towards a higher glycolytic index. We performed an in vivo glucose uptake assay employing fluorescent derivative of glucose, 2-NBDG (Zou et al., 2005) in the late third instar lymph gland of both control and FAO mutant. In control, late third instar Dome+ progenitors exhibit low glucose uptake (Figure 5B–B'' and Figure 5D) compared to the higher uptake detectable in the peripheral hemocyte population marked by Hml (Figure 5—figure supplement 1A–A''' and quantified in Figure 5—figure supplement 1B). In sharp contrast, higher glucose uptake is evident in the FAO mutant progenitors (Figure 5C–C'' and quantified in Figure 5D). In concordance with the above result, in vivo lactate dehydrogenase assay (Abu-Shumays and Fristrom, 1997) also revealed a high glycolytic index prevalent in the lymph glands of homozygous whd1 (Figure 5E–F).

Figure 5 with 1 supplement see all
FAO loss in hemocyte progenitors led to sustained glycolysis.

(A) ATP levels in control and whd1/whd1 whole larvae. p-Value of whd1/whd1compared to control = 5.327×10−2. (B–D) Glucose incorporation (marked by 2-NBDG uptake) levels in control dome-MESO-EBFP2/+ (B–B'') and dome-MESO-EBFP2/+; whd1/whd1 (C–C'') lymph glands. (D). Quantitative analysis of results from B–C demonstrating a significant increase in glucose uptake in the whd1/whd1 progenitors. p-Value for dome-MESO-EBFP2/+; whd1/whd1 = 6.09×10-7 compared to control. (E–F) Increased lactate dehydrogenase in-vivo enzymatic staining assay of whd1/whd1 lymph gland (F) compared to control (E). (G) Fold change in the level of Hex-A mRNA expression in control w1118 and whd1/whd1 lymph glands. p-Value of whd1/whd1 = 6.379×10−3 compared to control. (H) Fold change in the level of Pfk mRNA expression in control w1118 and whd1/whd1 lymph glands. p-Value of whd1/whd1 = 3.739×10−3 compared to control. (I–M) Proliferation status (marked by EdU) in control dome > GFP (I–I'), dome > GFP; whd1/whd1 (J–J'), 2-DG fed dome > GFP; whd1/whd1 (K–K') and dome > GFP; whd1/whd1; UAS-Glut1 RNAi (L–L') lymph glands. (M). Quantitative analysis of results from I–L'. p-Value for dome > GFP; whd1/whd1 = 4.37×10−7 compared to control and p-value for dome > GFP; whd1/whd1 = 3.25×10−7 fed with 2-DG compared to non-fed dome > GFP; whd1/whd1. p-Value for dome > GFP; whd1/whd1; UAS-Glut1 RNAi = 4.53×10−7 compared to non-fed dome > GFP; whd1/whd1. (M–P) Comparison of differentiation (marked by P1) levels in control dome > GFP (M), dome > GFP; whd1/whd1 (N) and 2-DG fed dome > GFP; whd1/whd1 (O) lymph glands. (P). Quantitative analysis of results from M–O show decline in proliferation upon 2-DG feeding. p-Value for dome > GFP; whd1/whd1 = 4.43×10−11 compared to control and p-Value for dome > GFP; whd1/whd1 = 8.6×10−2 fed with 2-DG compared to non-fed dome > GFP; whd1/whd1. p-Value for dome > GFP; whd1/whd1; UAS-Glut1 RNAi = 5.9×10−2 compared to non-fed dome > GFP; whd1/whd1. n.s. = not significant. Individual dots represent biological replicates. Values are mean ± SD, asterisks mark statistically significant differences (*p<0.05; **p<0.01; ***p<0.001, Student’s t-test). Scale bar: 20 µm.

Moreover, in the whd1 lymph gland, the transcript levels of HexA (Hexokinase A) and Pfk (Phosphofructokinase), the enzymes involved in the two irreversible steps of glycolysis: exhibit a 1.6 fold and 1.7 fold increase in their expression, respectively (Figure 5G–H). Together these results reveal that upon FAO disruption, lymph gland progenitors adopt high-glucose utilization/metabolism.

Based on the above observations, we inferred that higher proliferation and differentiation defects observed in hemocyte progenitors with compromised FAO might be due to the surge in glucose uptake/metabolism. Upon rearing whd1 homozygous larvae in food supplemented with glycolysis inhibitor 2-Deoxy-D-glucose (2-DG), the otherwise hyper-proliferating hemocyte progenitors (EdU, Figure 5J) demonstrate a significant drop in EdU incorporation to a level that is comparable to control (compare Figure 5K with Figure 5I, and quantified in Figure 5M). Although the glycolytic block by feeding 2-DG rescues the cell cycle status, the abrogated differentiation observed in homozygous whd1 hemocyte progenitors is not restored (compare Figure 5P with Figure 5O and N and quantified in Figure 5R). Inhibition of glucose uptake by genetic perturbation of glucose transporter Glut1 in progenitor specific manner in the FAO mutant also endorses the above result (Figure 5L–L' and M, and Figure 5Q–R).

From these observations, it is evident that the surge in glucose metabolism encountered by the hemocyte progenitors upon FAO loss is responsible for their altered cell cycle. However, the glycolytic surge upon FAO disruption is unable to initiate progenitor differentiation. Collectively, the above results indicate that FAO plays a critical role in regulating the differentiation of hemocyte progenitors.

FAO loss in progenitors causes an altered histone acetylation

Acetyl-CoA generated from FAO, apart from serving as a substrate for the Krebs cycle, is essential for the acetylation of various proteins, including histones. We, therefore, wondered that disruption of FAO in whd1 hemocyte progenitors might also result in altered histone acetylations, which may, in turn, result in cell cycle and differentiation defects.

Histone acetylation mediated by Histone Acetyl Transferases (HATs) directly controls the expression of differentiation factor, thereby regulating germline stem cell differentiation (McCarthy et al., 2018; Xin et al., 2013). To ascertain whether HATs play a similar role in hemocyte progenitor differentiation, progenitor-specific RNAi-mediated knockdown of HATs function was done following the timeline, as shown in Figure 2—figure supplement 1D. Quite strikingly, loss of Histone Acetyl Transferase (HAT) genes, Gcn5 (Carré et al., 2005) and chm (chameau) (Grienenberger et al., 2002; Miotto et al., 2006) phenocopy the differentiation defect seen in the hemocyte progenitors of FAO loss of function (Figure 6A–C). Additionally, downregulation of Acetyl Coenzyme A synthase/AcCoAS (the Drosophila orthologue of ACSS2) (Mews et al., 2017) results in a phenotype identical to HAT or FAO loss (Figure 6D) confirming the essential role of acetylation in hemocyte progenitor differentiation.

Figure 6 with 1 supplement see all
Hemocyte progenitors of HAT and FAO loss of function exhibits altered histone acetylation.

(A–H) Comparison of differentiation (marked by P1) levels in dome > GFP lymph gland of control (A) with progenitor-specific downregulation of (B) chm, (C) Gcn5, (D) AcCoAS and (E) whd1/whd1, (F) transheterozygote of whd and ATPCL (whd1/ATPCL01466) and transheterozygote of whd and sea (whd1/seaEPEP3364) (G). (H) Quantitative analyses of the results from A–G. p-Value for dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-chm RNAi = 2.267×10−15 compared to control. p-Value for dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-Gcn5 RNAi = 1.990×10−14 compared to control. p-Value for dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-AcCoAS RNAi = 2.601×10−15 compared to control. p-Value for dome-GAL4, UAS-GFP; whd1/whd1 = 1.400×10−15 compared to control. p-Value for dome-GAL4, UAS-GFP; whd1/ATPCL01466 = 6.835×10−3 compared to control dome-GAL4, UAS-GFP; whd1/whd1. p-Value for dome-GAL4, UAS-GFP; whd1/seaEP3364 = 2.974×10−2 compared to control dome-GAL4, UAS-GFP; whd1/whd1. (I–J) Western blot analysis of H3K9 acetylation level in control OreR and whd1/whd1 larvae with H3 as a loading control (I). (J). Quantitative analysis of H3K9 acetylation level in I. p-Value for whd1/whd1 = 8.056×10−3 compared to control OreR. (K–P) Clonal analysis of histone acetylation in the GFP-positive hs-Flp/Ay-GAL4 based clonal patches (GFP indicates cells where the whd function is knocked down). Immunostaining with H3 (K–L), H3K9 acetylation (M–N), and H4 pan acetylation (O–P) antibodies. (L). Quantitative analyses of H3 acetylation level in K–K'''. p-Value for hs-Flp/Ay-GAL4. UAS-GFP; UAS-whd RNAi = 8.188×10−1 compared to control. (N). Quantitative analysis of H3K9 acetylation level in (M–M'''). p-Value for hsFlp/+; Ay-GAL4. UAS-GFP; UAS-whd RNAi = 2.238×10−12 compared to control. (P). Quantitative analysis of H4 acetylation level in O–O'''. p-Value for hsFlp/Ay-GAL4. UAS-GFP, UAS-whd RNAi = 1.083×10−9 compared to control. Individual dots represent biological replicates. Values are mean ± SD, asterisks mark statistically significant differences (*p<0.05; **p<0.01; ***p<0.001, Student’s t-test). Scale bar: 20 µm n.s. = not significant.

Next, we performed an epistatic interaction of whd1 allele with other major acetyl-CoA related genes. Citrate transporter SLC25A1 or scheggia/sea in Drosophila (Carrisi et al., 2008) exports Krebs cycle metabolite citrate from the mitochondria to the cytoplasm to generate acetyl-CoA. In the cytosol, citrate gets converted to acetyl-CoA by ATP citrate lyase (ACLY encoded by Drosophila orthologue, ATPCL). Trans-heterozygous loss-of-function allelic combinations of whd1 with either ATPCL01466 (Figure 6F) or seaEP3364 (Figure 6G) phenocopy differentiation defects seen in whd1 homozygous lymph gland (Figure 6E). Above set of genetic correlations reveal that alteration in histone acetylation affects hemocyte progenitor differentiation (quantified in Figure 6H).

Knockdown of the rate-limiting enzyme of FAO:CPT1, leads to a reduced level of H3K9 acetylation in lymphatic endothelial cells (Wong et al., 2017). We wondered whether similar acetylation defect occurs in whd1 homozygous mutant larvae. Immunoblot analysis of extracted histones with antibody against acetylated anti-H3K9 reveals that whd1 homozygous larvae have low levels of H3K9 acetylation (Figure 6I–J) when compared to control. However, the level of histone H3 is comparable to that of control.

Whether the tissue of our interest also reflects this decline in acetylation level, H3K9 acetylation labeling was performed in mosaic clones using hsFlp/Ay-GAL4 mediating RNAi knockdown of Drosophila CPT1 orthologue whd. The clonal patches positively marked with GFP (where whd has been downregulated) show a significant drop in H3K9 acetylation levels (Figure 6M–N) compared to surrounding hemocyte progenitors. However, histone H3 labeling in both mutant and control clonal patches are comparable (Figure 6K–L) and serve as a control. This observation, along with the western blot analyses, reveals the occurrence of H3K9 acetylation defects in FAO loss of function. Likewise, histone H4 acetylation visualized by pan anti-H4 acetylation antibody reveals a drastic drop in whd knockdown clonal patches (Figure 6O–P). Both the expression of H3K9 and pan H4 acetylation remains unaltered in mock/wild type clones (Figure 6—figure supplement 1A–D). Further, upon progenitor specific downregulation of whd function, a decline in the level of both H3K9 acetylation (compare Figure 6—figure supplement 1H–H' with Figure 6—figure supplement 1I–I' and J) and pan H4 acetylation (Figure 6—figure supplement 1K–K’ with Figure 6—figure supplement 1L–L' and M) is evident. In all these scenarios, histone H3 labeling remains unaffected (Figure 6—figure supplement 1E–G).

Above molecular and genetic analyses demonstrate that the downregulation of FAO in the hemocyte progenitors leads to a decline in histone acetylation. The next step was to correlate whether the differentiation defects of hemocyte progenitors in FAO loss of function is a consequence of altered histone acetylation levels.

Acetate supplementation rescues differentiation defects of FAO mutant hemocyte progenitors

Histone acetylation in eukaryotes relies on acetyl-Coenzyme A (acetyl-CoA). It has been established earlier that compromised histone acetylation levels can be restored by supplementing acetate (Carrisi et al., 2008; Wellen et al., 2009; Wong et al., 2017). The supplemented acetate is converted to acetyl-CoA, which restores the endogenous histone acetylation in a cell. We wondered whether replenishing the H3K9 acetylation levels in whd1 by acetate supplementation (50 mM, supplemented fly food post first instar) can rescue the differentiation defect seen in those lymph glands.

Intriguingly, acetate feeding does not affect progenitor differentiation in control, whereas it rescues the differentiation defects seen in homozygous whd1 hemocyte progenitors (Figure 7A–E). At the molecular level, we observed that acetate supplementation to whd1 mutant larvae leads to a restoration of the H3K9 acetylation level (Figure 7F–G), which might lead to the rescue of the differentiation defect (Figure 7A–E). In order to probe this possibility, the lymph gland from acetate fed larvae were dissected and assayed for the status of H3K9 acetylation level. Figure 7H–L reveals that acetate supplementation indeed restores the compromised acetylation status in the whd1 lymph gland (compare Figure 7J–J' with Figure 7K–K').

Acetate supplementation rescues differentiation defects of FAO mutant hemocyte progenitors.

(A–E) Comparison of differentiation (marked by P1) levels in dome > GFP lymph gland of control (A) dome > GFP supplemented with acetate (B) dome > GFP; whd1/whd1 (C) and dome > GFP; whd1/whd1 supplemented with acetate (D). (E). Quantitative analysis of results from A–D. p-Value for dome-GAL4, UAS-GFP; tubGAL80ts20 = 2.718×10-1 fed with acetate compared to control dome-GAL4, UAS-GFP; tubGAL80ts20. p-Value for dome-GAL4, UAS-GFP; whd1/whd1 = 3.18×10−16 compared to control dome-GAL4, UAS-GFP; tubgal80ts20. p-Value for dome-GAL4, UAS-GFP; whd1/whd1 = 2.576×10−10 fed with acetate compared to control dome-GAL4, UAS-GFP; tubGAL80ts20. (F–G) Western blot analysis of H3K9 acetylation levels in control OreR and whd1/whd1 larvae supplemented with acetate and non-fed controls with H3 as a loading control. (G) Quantitative analysis of H3K9 acetylation levels in F. p-Value for OreR = 4.589×10−3 supplemented with acetate compared to non-fed control OreR. p-Value for non-fed whd1/whd1 = 8.001×10−3 compared to non-fed control OreR. p-Value for acetate supplemented whd1/whd1 = 3.582×10−2 compared to non-fed control whd1/whd1. (H–L) Acetate supplementation restores H3K9 acetylation status in the whd1/whd1 lymph gland (H-Iʹ). (L) Quantitative analysis of acetylation level in control, whd mutant, and whd mutant fed on acetate. p-Value for acetate supplemented dome-GAL4, UAS-GFP = 1.38×10−1 compared to non-fed control. p-Value for dome-GAL4, UAS-GFP; whd1/whd1 = 1.276×10−7 compared to dome-GAL4, UAS-GFP. p-Value for acetate supplemented dome-GAL4, UAS-GFP; whd1/whd1 = 1.31×10−6 compared to non-fed control dome-GAL4; UAS-GFP; whd1/whd1. (M–O) Comparison of H3K9 acetylation level in Dome+ progenitors of L-carnitine fed larvae (N–N') with non-fed control (M–M'). (O) Quantitative analysis of H3K9 acetylation levels in M–N'. p-Value for dome-GAL4, UAS-GFP = 1.079×10−8 supplemented with L-carnitine compared to non-fed control dome-GAL4, UAS-GFP. (P–Q) Western blot analysis of H3K9 acetylation levels in OreR larvae supplemented with L-carnitine and non-fed controls with H3 as a loading control. Quantitative analysis of H3K9 acetylation levels in N. p-Value for OreR = 3.17×10−4 supplemented with L-carnitine compared to non-fed control OreR. Individual dots represent biological replicates. Values are mean ± SD, asterisks mark statistically significant differences (*p<0.05; **p<0.01; ***p<0.001, Student’s t-test). Scale bar: 20 µm n.s. = not significant.

Conversely, upregulation of FAO by L-carnitine feeding leads to elevated H3K9 acetylation in the lymph gland (compare Figure 7M–M' with Figure 7N–N' and quantified in Figure 7O). Likewise, the level of H3K9 acetylation in L-carnitine fed larvae demonstrates a significant upregulation when compared to age-isogenised non-fed control larvae (Figure 7P–Q).

These results establish that the hemocyte progenitors require FAO mediated histone acetylation for their differentiation.

JNK signaling regulates FAO in the hemocyte progenitors

Next, we attempted to understand how the FAO-mediated metabolic circuitry collaborates with the known differentiation signals of hemocyte progenitors. Jun-Kinase and dFOXO (Forkhead box O) mediated signal has been previously implicated for hemocyte progenitor differentiation (Owusu-Ansah and Banerjee, 2009). Analogous to whd1 lymph glands, expression of a dominant-negative allele of basket (bsk, Drosophila orthologue of Jun-Kinase) in the progenitors results in stalled differentiation (Figure 8A–C). On the other hand, overexpression of FOXO, pushes the progenitor fate towards precocious differentiation (Owusu-Ansah and Banerjee, 2009; Figure 8D–E and Figure 8H). However, genetic removal of one copy of whd is sufficient enough to prevent the precocious differentiation as observed in progenitor-specific overexpression of FOXO (Figure 8F–H). These results illustrate an unappreciated link between the differentiation signals and FAO in the lymph gland progenitors.

Figure 8 with 1 supplement see all
JNK regulates FAO in hemocyte progenitors of larval lymph gland.

(A–C) Comparison of differentiation (marked by P1) levels in dome > GFP lymph gland of control (A), and bsk/JNK knockdown in hemocyte progenitors by dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-bskDN (B). (C). Quantitative analysis of the differentiation level from A–B reveals a significant increase in the Dome+ progenitor zone and a decrease in differentiation. p-Value for dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-bskDN=5.84×10−11 compared to control. (D–H) Differentiation levels (red, marked by Pxn) in overexpression of FOXO by dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-FOXO (E) is significantly increased compared to control (D). The increased differentiation in FOXO overexpression background is significantly rescued by one copy of the null allele of whd (G). (F). The differentiation level in one copy null allele of whd. (H). Quantitative analysis of the differentiation level from D–G reveals a significant increment in Pxn+ differentiated cell area in FOXO overexpression from Dome+ progenitors, which is significantly rescued by one copy null allele of whd. p-Value for dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-FOXO =5.77×10−11 compared to control. p-Value for dome-GAL4, UAS-GFP; whd1/+ = 2.11×10−5 compared to control. p-Value for dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-FOXO/whd1 = 3.84×10−9 compared to dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-FOXO. (I) Real-time expression analysis of fatty acid oxidation enzymes, whd, Mcad, Mtpα, scully, Mtpβ, and yip2 from dome > GFP and dome > GFP > UAS-bskDN lymph glands. The expression of whd shows a significant drop ~41% in dome > GFP > UAS-bskDN compared to control dome > GFP. p-Value for whd expression in dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-bskDN=7.06×10−3 compared to control. p-Value for Mcad expression in dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-bskDN=6.71×10−1 compared to control. p-Value for Mtpα expression in dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-bskDN=8.95×10−1 compared to control. p-Value for scully expression in dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-bskDN=9.73×10−1 compared to control. p-Value for Mtpβ expression in dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-bskDN=7.7×10−1 compared to control. p-Value for yip2 expression in dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-bskDN=2.42×10−2 compared to control. (J–O) Clonal analysis of histone acetylation in GFP-positive hsFlp/Ay-GAL4 based clonal patches expressing a dominant-negative form of bsk and immunostaining with H3 (J–J'''), H3K9 acetylation (L–L''') and H4 pan acetylation (N–N''') antibodies. (K). Quantitative analysis of H3 acetylation level in J–J'''. p-Value for hsFlp/Ay-GAL4. UAS-GFP, UAS-bskDN = 6.32×10−1 compared to control. (M). Quantitative analysis of H3K9 acetylation level in L–L'''. p-Value for hsFlp/Ay-GAL4. UAS-GFP; UAS-bskDN = 1.911×10−7 compared to control. (O). Quantitative analysis of H4 acetylation level in N–N'''. p-Value for hs-Flp/Ay-GAL4. UAS-GFP, UAS-bskDN = 8.22×10−9 compared to control. (P–T) Stalled differentiation levels (red, marked by Pxn) in dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-bskDN (R) is significantly rescued in larvae reared in fly food supplemented with acetate (S). The differentiation level in control (P) dome-GAL4, UAS-GFP; tubGAL80ts20 remain unaltered upon acetate feeding (Q). (T). Quantitative analysis of the differentiation level from P–S reveals a significant rescue of differentiated cells upon acetate supplementation in dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-bskDN lymph glands. p-Value for acetate supplemented dome-GAL4, UAS-GFP; tubGAL80ts20 = 5.655×10−1 compared to non-fed control. p-Value for dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-bskDN=1.32×10−8 compared to control dome-GAL4, UAS-GFP; tubGAL80ts20. p-Value for acetate fed dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-bskDN=4.73×10−7 compared to non-fed dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS bskDN. (U–Y) The differentiation level (red, marked by Pxn) in control (U) dome-GAL4, UAS-GFP; tubGAL80ts20 increases upon L-carnitine feeding (V). Defect in differentiation levels (red, marked by Pxn) in dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS bskDN (W) is significantly rescued in larvae reared in fly food supplemented with L-carnitine (X). (Y). Quantitative analysis of the differentiation level from U–X, reveals a significant rescue of differentiated cells upon L-carnitine supplementation in dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-bskDN lymph glands. p-Value for L-carnitine supplemented dome-GAL4, UAS-GFP; tubGAL80ts20 = 1.69×10−8 compared to non-fed control. p-Value for dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-bskDN =4.5×10−10 compared to control dome-GAL4, UAS-GFP; tubGAL80ts20. p-Value for L-carnitine fed dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS-bskDN =8.307×10−9 compared to non-fed dome-GAL4, UAS-GFP; tubGAL80ts20 > UAS -bskDN. Individual dots represent biological replicates. Values are mean ± SD, asterisks mark statistically significant differences (*p<0.05; **p<0.01; ***p<0.001, Student’s t-test). Scale bar: 20 µm.

Next, we addressed whether JNK signaling regulates the expression of genes of FAO. The transcription of CPT1/whd, Mcad, Mtpα, scully, Mtpβ, and yip2 was assayed upon down-regulation of bsk from hemocyte progenitors. The transcript level of CPT1/whd indicates a ~ 41% drop (Figure 8I) while the expression of rest of the enzymes either exhibited a mild drop (~18% in yip2) or no significant alteration upon loss of bsk from the progenitors. This observation established that JNK controls the transcription of CPT1/whd, thereby regulating FAO.

Interestingly, clonal analyses of bsk/JNK knockdown in hemocyte progenitors reveals a drop in the levels of H3K9 acetylation (Figure 8L–M and Figure 8—figure supplement 1A–A''') and H4 pan acetylation (Figure 8N–O and Figure 8—figure supplement 1B–B''') similar to whd downregulated clonal patches (Figure 6M–P). Additionally, expression of bskDN in progenitor-specific manner brings about a conspicuous downregulation of both H3K9 acetylation (Figure 8—figure supplement 1F–H) and H4 pan acetylation (Figure 8—figure supplement 1I–K). However, in the above experiments, histone H3 labeling remains unaffected (Figure 8J–K and Figure 8—figure supplement 1C–E).

Since compromised acetylation leads to differentiation defect in whd1, which can be rescued upon acetate feeding, we wondered whether the block in differentiation upon JNK loss could be rescued similarly. Indeed, the hemocyte progenitors that lacked JNK (dome >bskDN) when reared in acetate supplemented food, demonstrate differentiation levels comparable to the similarly-aged control (Figure 8P–T).

Therefore, lack of histone acetylation encountered in JNK loss leads to the block in hemocyte progenitor differentiation. Moreover, whd transcription is under the regulation of JNK, which further endorses that JNK mediated regulation of FAO is crucial for differentiation.

If this is true, then the upregulation of FAO in JNK loss either by L-carnitine supplementation or by overexpression of whd should facilitate differentiation. Our results demonstrate that the upregulation of FAO in hemocyte progenitors that lacked JNK indeed elicits differentiation (Figure 8U–Y and Figure 8—figure supplement 1L–P).

Collectively, these results are in agreement with the fact that JNK regulates the differentiation of hemocyte progenitors by FAO-mediated histone acetylation.

Discussion

Along with cellular signaling network, stem/progenitor cell fate and state are directly governed by their metabolism in normal development and during pathophysiological conditions (Ito and Ito, 2016; Ito and Suda, 2014; Oginuma et al., 2017; Shyh-Chang et al., 2013; Shyh-Chang and Ng, 2017). How the metabolic circuitry works in sync with the cell signaling machinery to achieve cellular homeostasis is yet to be fully understood. Here, we show the developmental requirement of FAO in regulating the differentiation of hemocyte progenitors in Drosophila.

Our molecular genetic analyses reveal a signaling cascade that links ROS-JNK-FAO and histone modification essential for the differentiation of hemocyte progenitors (Figure 9). High ROS levels in the progenitors evoke differentiation program by triggering JNK and FOXO mediated signals (Owusu-Ansah and Banerjee, 2009). We show that activated JNK, in turn, leads to transcriptional induction of whd to facilitate the import of fat moiety into mitochondria for β-oxidation. Optimal level of acetyl-CoA, the end product of FAO, is critical for the acetylation of several proteins, including histones. Altering this pathway either at the level of JNK or FAO affects histone acetylation in a HAT dependent manner. On the other hand, it is quite possible that precocious differentiation of blood progenitors in the lymph gland of starved larvae (Shim et al., 2012) might be an outcome of starvation induced fat mobilization and increased FAO.

The regulation of FAO by JNK is critical for differentiation.

ROS-JNK link has been previously shown to be essential for differentiation (Owusu-Ansah and Banerjee, 2009). The G2-M arrested hemocyte progenitors employ β-oxidation for their differentiation. ROS–JNK circuit impinges on FAO to facilitate progenitor differentiation. JNK signaling transcriptionally regulates whd, the rate-limiting enzyme of FAO leading to the production of acetyl-CoA. Acetyl-CoA leads to acetylation of histones in the hemocyte progenitors, which is critical for their differentiation.

JNK signaling has been associated with histone acetylation in different biological processes (Miotto et al., 2006; Wu et al., 2008). In Drosophila, Fos, a transcriptional activator of JNK, interacts with Chm (HAT) and causes modification of histones. Our investigation reveals that indeed upon downregulation of JNK (Figure 8A–C) and Chm (Figure 6A–B), differentiation of hemocyte progenitors is halted. Further, we show that halt in differentiation upon JNK loss is associated with alteration of the acetylation profile of H3K9 and H4. Given the fact that JNK signaling regulates FAO, which in turn provides the acetyl moiety for histone acetylation, our study provides a new dimension in JNK’s role for histone acetylation in an FAO dependent manner. As a result, despite having high ROS levels, the hemocytes fail to differentiate if FAO is attenuated. Thus, the current work provides a metabolic link between JNK and epigenetic regulation of gene expression.

Our results show that upon disruption of FAO, hematopoietic progenitors adopt glycolysis to overcome the G2-M arrest but fail to initiate differentiation. Pharmacological and genetic inhibition of glycolysis in the FAO mutant restores their cell cycle defect but fails to facilitate their differentiation. The glycolytic surge in FAO mutants is not capable to take them through the differentiation process. This indicates that for the process of differentiation, the acetyl moiety derived from FAO plays a key role to facilitate hemocyte progenitor differentiation.

A recent study has demonstrated that alteration in acetyl-CoA levels can affect proteome and cellular metabolism by modulating intracellular crosstalk (Dieterich et al., 2019). It is intriguing to see that restoring acetylation level by the acetate supplementation is capable of rescuing hematopoietic defects in the lymph gland progenitors of FAO mutants. The acetate supplemented is converted into the end product of FAO: acetyl-CoA, the metabolite that is essential for histone acetylation. The involvement of acetyl-CoA in facilitating differentiation is further evidenced when on genetically downregulating AcCoAs (the major enzyme in acetyl-CoA generation) from the progenitor leads to a halt in their differentiation. Our study thus establishes that for hemocyte progenitor differentiation, the metabolic process FAO involves its metabolite acetyl-CoA for epigenetic modification. Earlier studies in diverse model systems have demonstrated that compromised in vivo histone acetylation defects can be rescued by acetate supplementation (Gao et al., 2016; Soliman et al., 2012). A similar finding in Drosophila hematopoiesis signifies the relevance of acetate supplementation across taxa. In light of this study, it would be interesting to see whether metabolite supplementation of FAO can modulate pathophysiological scenarios like certain forms of cancer which rely on fat oxidation.

FAO has been implicated in HSCs maintenance downstream of the PML-Peroxisome proliferator-activated receptor delta (PPARδ) pathway (Ito et al., 2012). Mechanistically, the PML-PPARδ-FAO pathway regulates HSC maintenance by controling asymmetric division. In FAO inhibition, HCSs undergo symmetric divisions, which lead to exhaustion and depletion of the stem cell pool resulting in their differentiation (Ito et al., 2012). Another study in mice shows that upon short term starvation, there is a decline in the number of HSC (Takakuwa et al., 2019). Since the HSC maintenance is FAO dependent (Ito et al., 2012), a loss in number might be attributed to heightened fat oxidation during starvation. Interestingly, metabolic dependence on FAO has been reported in mammalian neural stem cell (Knobloch et al., 2017), muscle stem cells (Ryall et al., 2015) and intestinal stem cells (Chen et al., 2020). Although endothelial precursors (Wong et al., 2017) is also known to be dependent on FAO, it remains to be seen whether FAO is a preferred metabolic requirement for progenitor differentiation.

The entire blood cell repertoire in Drosophila is engaged in innate immunity, maintenance of tissue integrity, wound healing, and heterogeneous stress responses, and is therefore functionally considered to be similar to myeloid cells in mammals (Banerjee et al., 2019; Gold and Brückner, 2014). Interestingly, several molecular mechanisms that regulate Drosophila lymph gland hematopoiesis are essential players in progenitor-based hematopoiesis in vertebrates (Banerjee et al., 2019; Gold and Brückner, 2014; Krzemien et al., 2010).

Based on the above conservations, it is reasonable to propose the requirement of FAO in progenitor differentiation described here will help us in understanding mammalian myeloid progenitor differentiation.

Materials and methods

Key resources table
Reagent type
(species) or resource
DesignationSource or referenceIdentifiersAdditional
information
Gene (Drosophila melanogaster)domeFlybase:FB2020_01FLYB:FBgn
0043903
Gene (Drosophila melanogaster)HmlFlybase:FB2020_01FLYB:FBgn
0029167
Gene (Drosophila melanogaster)Tep4Flybase:FB2020_01FLYB:FBgn
0031888
Gene (Drosophila melanogaster)CG3902Flybase:FB2020_01FLYB:FBgn
0036824
Gene (Drosophila melanogaster)MtpαFlybase:FB2020_01FLYB:FBgn
0041180
Gene (Drosophila melanogaster)MtpβFlybase:FB2020_01FLYB:FBgn
0025352
Gene (Drosophila melanogaster)whdFlybase:FB2020_01FLYB:FBgn
0261862
Gene (Drosophila melanogaster)Hnf4Flybase:FB2020_01FLYB:FBgn
0041180
Gene (Drosophila melanogaster)chmFlybase:FB2020_01FLYB:FBgn
0028387
Gene (Drosophila melanogaster)Gcn5Flybase:FB2020_01FLYB:FBgn
0020388
Gene (Drosophila melanogaster)AcCoASFlybase:FB2020_01FLYB:FBgn
0012034
Gene (Drosophila melanogaster)Glut1Flybase:FB2020_01FLYB:FBgn
0264574
Gene (Drosophila melanogaster)ATPCLFlybase:FB2020_01FLYB:FBgn
0020236
Gene (Drosophila melanogaster)seaFlybase:FB2020_01FLYB:FBgn
0037912
Gene (Drosophila melanogaster)bskFlybase:FB2020_01FLYB:FBgn
0000229
Genetic reagent(Drosophila melanogaster)dome-GAL4BloomingtonDrosophilaStock CenterBDSC:81010; FLYB:FBti0022298; RRID:BDSC_81010FlyBase symbol: P{GawB}domePG14
Genetic reagent (Drosophila melanogaster)Hml-dsRed.ΔMakhijani et al., 2011FLYB:FBgn
0041180
FlyBase symbol: P{Hml-dsRed.Δ}
Genetic reagent (Drosophila melanogaster)HmlΔ-GAL4Sinenko and Mathey-Prevot, 2004FLYB: FBgn
0040877
FlyBase symbol: P{Hml-GAL4.Δ}
Genetic reagent (Drosophila melanogaster)Pvf2-lacZChoi et al., 2008FLYB:FBtp0052107FlyBase symbol: P{Pvf2-lacZ.C}
Genetic reagent (Drosophila melanogaster)TepIV-GAL4Kyoto Stock CenterDGGR:105442;
FLYB:FBti0037434;
RRID:DGGR_105442
FlyBase symbol: P{GawB}NP7379
Genetic reagent (Drosophila melanogaster)CG3902-YFPKyoto Stock CenterDGGR:115356;
FLYB:FBti0143519;
RRID:DGGR_115356
FlyBase symbol: PBac{566 .P.SVS-1}CG3902CPTI100004
Genetic reagent (Drosophila melanogaster)Mtpα[KO]Kyoto Stock CenterDGGR:116261;
FLYB:FBal0267653;
RRID:DGGR_116261
FlyBase symbol: MtpαKO
Genetic reagent (Drosophila melanogaster)Mtpβ[KO]Kyoto Stock CenterDGGR:116262;
FLYB:FBal0267654;
RRID:DGGR_116262
FlyBase symbol: MtpβKO
Genetic reagent (Drosophila melanogaster)UAS-whd RNAi [KK]ViennaDrosophila RNAi CenterVDRC:v105400;
FLYB:FBti0116709;
RRID:FlyBase_FBst0477227
FlyBase symbol: P{KK100935}VIE-260B
Genetic reagent (Drosophila melanogaster)OreRBloomingtonDrosophilaStock CenterBDSC:5; FLYB:FBsn0000277; RRID:BDSC_5FlyBase symbol: Oregon-R-C
Genetic reagent (Drosophila melanogaster)w[1118]BloomingtonDrosophilaStock CenterBDSC:3605; FLYB:FBal0018186;RRID:BDSC_3605FlyBase symbol: w1118
Genetic reagent (Drosophila melanogaster)UAS-Hnf4.miRNABloomingtonDrosophilaStock CenterBDSC:44398; FLYB:FBti0152533;RRID:BDSC_44398FlyBase symbol: P{UAS-Hnf4.miRNA}attP16
Genetic reagent (Drosophila melanogaster)UAS-whd RNAiBloomingtonDrosophilaStock CenterBDSC:34066; FLYB:FBal0263076; RRID:BDSC_34066FlyBase symbol: whdHMS00040
Genetic reagent (Drosophila melanogaster)UAS-FOXO.PBloomingtonDrosophilaStock CenterBDSC:9575; FLYB:FBtp0017636; RRID:BDSC_9575FlyBase symbol: P{UAS-foxo.P}
Genetic reagent (Drosophila melanogaster)Hnf4-GAL4BloomingtonDrosophilaStock CenterBDSC:47618; FLYB:FBti0136396; RRID:BDSC_47618FlyBase symbol: P{GMR50A12-GAL4}attP2
Genetic reagent (Drosophila melanogaster)UAS-FUCCIBloomingtonDrosophilaStock CenterBDSC:55121; RRID:BDSC_55121FlyBase symbol: P{UAS-GFP.E2f1.1–230}32; P{UAS-mRFP1.NLS.CycB.1–266}19
Genetic reagent (Drosophila melanogaster)UAS-mito-HA-GFPBloomingtonDrosophilaStock CenterBDSC:8442; FLYB:FBti0040803; RRID:BDSC_8442FlyBase symbol: P{UAS-mito-HA-GFP.AP}2
Genetic reagent (Drosophila melanogaster)UAS-chm RNAiBloomingtonDrosophilaStock CenterBDSC:27027; FLYB:FBal0220716; RRID:BDSC_27027FlyBase symbol: chmJF02348
Genetic reagent (Drosophila melanogaster)UAS-Gcn5 RNAiBloomingtonDrosophilaStock CenterBDSC:33981; FLYB:FBal0257611; RRID:BDSC_33981FlyBase symbol: Gcn5HMS00941
Genetic reagent (Drosophila melanogaster)UAS-AcCoAS RNAiBloomingtonDrosophilaStock CenterBDSC:41917; FLYB:FBal0279313; RRID:BDSC_41917FlyBase symbol: AcCoASHMS02314
Genetic reagent (Drosophila melanogaster)UAS-Glut1RNAiBloomingtonDrosophilaStock CenterBDSC:28645; FLYB:FBal0239561; RRID:BDSC_28645FlyBase symbol: Glut1JF03060
Genetic reagent (Drosophila melanogaster)ATPCL[01466]BloomingtonDrosophilaStock CenterBDSC:11055; FLYB:FBal0007976; RRID:BDSC_11055FlyBase symbol: ATPCL01466
Genetic reagent (Drosophila melanogaster)sea[EP3364]BloomingtonDrosophilaStock CenterBDSC:17118; FLYB:FBal0131420; RRID:BDSC_17118FlyBase symbol: seaEP3364
Genetic reagent (Drosophila melanogaster)UAS-bsk[DN]BloomingtonDrosophilaStock CenterBDSC:6409;
FLYB:FBti0021048; RRID:BDSC_6409
FlyBase symbol: P{UAS-bsk.DN}2
Genetic reagent (Drosophila melanogaster)UAS-mCD8::GFPBloomingtonDrosophilaStock CenterBDSC:5137; FLYB:FBti0180511; RRID:BDSC_5137FlyBase symbol: P{UAS-mCD8::GFP.L}2
Genetic reagent (Drosophila melanogaster)UAS-mCD8::RFPBloomingtonDrosophilaStock CenterBDSC:27400; FLYB:FBti0115747; RRID:BDSC_27400FlyBase symbol: P{UAS-mCD8.mRFP.LG}28a
Genetic reagent (Drosophila melanogaster)U-6;sgRNA-whd-KOBloomingtonDrosophilaStock CenterBDSC:77066; FLYB:FBal0335953; RRID:BDSC_77066FlyBase symbol: whdTKO.GS00854
Genetic reagent (Drosophila melanogaster)U-6;sgRNA-whd-OEBloomingtonDrosophilaStock CenterBDSC:68139; FLYB:FBal0337690; RRID:BDSC_68139FlyBase symbol: whdTOE.GS00536
Genetic reagent (Drosophila melanogaster)whd[1]BloomingtonDrosophilaStock CenterBDSC:441; FLYB:FBal0018515; RRID:BDSC_441FlyBase symbol: whd1
Genetic reagent (Drosophila melanogaster)Hnf4[Δ33]BloomingtonDrosophilaStock CenterBDSC:43634; FLYB:FBal0240651; RRID:BDSC_43634FlyBase symbol: Hnf4Δ33
Genetic reagent (Drosophila melanogaster)Hnf4[Δ17]BloomingtonDrosophilaStock CenterBDSC:44218; FLYB:FBal0240650; RRID:BDSC_44218FlyBase symbol: Hnf4Δ17
Genetic reagent (Drosophila melanogaster)tubGAL80[ts20]BloomingtonDrosophilaStock CenterBDSC:7109; FLYB:FBti0027796; RRID:BDSC_7109FlyBase symbol: P{tubP-GAL80ts}20
Genetic reagent (Drosophila melanogaster)hsFlpBloomingtonDrosophilaStock CenterBDSC:1929; FLYB:FBti0000784; RRID:BDSC_1929FlyBase symbol: P{hsFLP}12
Genetic reagent (Drosophila melanogaster)Ay-GAL4, UAS-GFPBloomingtonDrosophilaStock CenterBDSC:4411; FLYB:FBti0012290;FBti0003040RRID:BDSC_4411FlyBase symbol: P{AyGAL4}25; P{UAS-GFP.S65T}Myo31DFT2
Antibodyanti-P1 (Mouse monoclonal)Kurucz et al., 2007Cat# NimC1, RRID:AB_2568423IF(1:50)
Antibodyanti-Pxn (Mouse)Nelson et al., 1994IF(1:400)
Antibodyanti-proPO (Rabbit polyclonal)Jiang et al., 1997IF(1:1000)
Antibodyanti-DE-cadherin (Rat polyclonal)Developmental Studies Hybridoma BankCat# DE-cad, RRID:AB_2314298IF(1:50)
Antibodyanti-Ci155(Rat polyclonal)Developmental Studies Hybridoma BankCat# 2A1,
RRID:AB_2109711
IF(1:2)
Antibodyanti-GFP (Rabbit polyclonal)InvitrogenCat# A-11122,
RRID:AB_221569
IF(1:100)
Antibodyanti-H3 (Rabbit polyclonal)Cell Signaling TechnologiesCat# 9927,
RRID:AB_330200
IF(1:400), WB(1:1000)
Antibodyanti-H3K9 acetylation (Rabbit polyclonal)Cell Signaling TechnologiesCat# 9927,
RRID:AB_330200
IF(1:300), WB(1:1000)
Antibodyanti-H4 pan acetylation (Rabbit polyclonal)Cell Signaling TechnologiesCat# 06–598,
RRID:AB_2295074
IF(1:500)
Chemical compound, drugSodium butyrateEMD Millipore19–137
Chemical compound, drugNicotinamideSigma-Aldrich72345
Chemical compound, drugEtomoxirCayman ChemicalsCay119695 µM
Chemical compound, drugMildronateCayman ChemicalsCay15997100 µM
Chemical compound, drugL-carnitine hydrochlorideSigma-AldrichC0283100 mM
Chemical compound, drug2-DGSigma-AldrichD8375100 mM
Chemical compound, drugSodium acetateSigma-Aldrich7119650 mM
Chemical compound, drug2-NBDGInvitrogenN131950.25 mM
Chemical compound, drugLipidTOXMolecular ProbesH344771:1000
Chemical compound, drugStreptavidin-Cy3Molecular ProbesSA10101:200
Chemical compound, drugNile redMolecular ProbesN11420.5 ug/mL
Chemical compound, drugDHE (Dihydroxy Ethidium)Molecular ProbesD113470.3 µM
Sequence-based reagentPfk_FThis paperPCR primersATCGTATTTTGGCTTGCCGC
Sequence-based reagentPfk_RThis paperPCR primersCCAGAGAGATGACCACTGGC
Sequence-based reagentHex_FThis paperPCR primersCTGCTTCTAACGGACGAACAG
Sequence-based reagentHex_RThis paperPCR primersGCCTTGGGATGTGTATCCTTGG
Sequence-based reagentwhd_FThis paperPCR primersGGCCAATGTGATTTCCCTGC
Sequence-based reagentwhd_RThis paperPCR primersTGCCCTGAACCATGATAGGC
Sequence-based reagentAct5C_FThis paperPCR primersACACATTTTGTAAGATTTGGTGTGT
Sequence-based reagentAct5C_RThis paperPCR primersCCGTTTGAGTTGTGCTGT
Sequence-based reagentMcad_FThis paperPCR primersGGCCTGGATCTCGATGTGTT
Sequence-based reagentMcad_RThis paperPCR primersGATCACAGGAGTTTGGCCCAG
Sequence-based reagentMtpα_FThis paperPCR primersATCACTGTTGGTGACGGACC
Sequence-based reagentMtpα_RThis paperPCR primersCTGCAGCAGTCTGATGGCTT
Sequence-based reagentscully_FThis paperPCR primersGATCAAGAACGCCGTTTCCC
Sequence-based reagentscully_RThis paperPCR primersCAGATCGGCCAGGATCACG
Sequence-based reagentMtpβ_FThis paperPCR primersCAGGCACTCGCTTTTGTCAT
Sequence-based reagentMtpβ_RThis paperPCR primersCCTGGCAATGTTGGAGGTCT
Sequence-based reagentyip2_FThis paperPCR primersTCTGCCGCAACCAAAGGTAT
Sequence-based reagentyip2_RThis paperPCR primersTTAAGACCGGCAGCATCCAG
Software, algorithmFijiFijiRRID:SCR_002285
Software, algorithmPhotoshop CCAdobeRRID:SCR_014199
Software, algorithmImarisBitplaneRRID:SCR_007370
Commercial assay or kitClick-iT EdU plus (DNA replication kit)InvitrogenC10639
Commercial assay or kitATP bioluminescence kit HSIISigma11699709001
Commercial assay or kitHistone extraction kitAbcamab113476
Commercial assay or kitRNAeasy Mini KitQiagen74104

Fly stocks

Request a detailed protocol

The fly stocks used were dome-GAL4, dome-MESO-EBFP2, Hml-DsRed (K. Bruckner), HmlΔ-GAL4 (S. Sinenko) Pvf2-LacZ (M. A. Yoo), TepIV-GAL4, CG3902-YFP, Mtpα[KO], Mtpβ[KO] (DGRC, Kyoto), UAS-whd RNAiKK (VDRC, Vienna), OreR, w1118, UAS-Hnf4.miRNA (Lin et al., 2009), UAS-whd RNAiHMS00040 (Manzo et al., 2018), UAS-FOXO.P, Hnf4-GAL4GMR50A12 (Tokusumi et al., 2017), UAS-FUCCI, UAS-mito-HA-GFP, UAS-chm-RNAiJF02348 (Dietz et al., 2015), UAS-Gcn5-RNAiHMS00941 (Janssens et al., 2017), UAS-AcCoAS RNAiHMS02314 (Eisenberg et al., 2014), UAS-Glut1 RNAiJF03060 (Charlton-Perkins et al., 2017), ATPCL01466, seaEP3364, UAS-bskDN, UAS-mCD8::GFP, U-6;sgRNA-whdTKO.GS00854, U-6;sgRNA-whdTOE.GS00536, whd1, Hnf4Δ33, Hnf4Δ17, tub-GAL80ts20, hsFlp and Ay-GAL4, UAS-GFP (BDSC, Bloomington Drosophila Stock Center).

Following genotypes were recombined for the current study:

  1. w; +/+; Hnf4-GAL4GMR50A12, UAS-mCD8::GFP

  2. dome-GAL4, UAS::mCD8RFP/FM7; +/+; +/+

  3. dome-GAL4, UAS-GFP/FM7; whd1/whd1; +/+

  4. w; U6:sgRNA-whd/U6:sgRNA-whd; UAS-dCas9/UAS-dCas9

  5. dome-GAL4/FM7; UAS-FUCCI/Cyo; +/+

  6. w; P{y[+t7.7] v[+t1.8]=TOE.GS00536}attP40/CyO; UAS-dCas9/UAS-dCas9

  7. dome-MESO-EBFP2/+; whd1/whd1; +/+

  8. hsFlp/hsFlp; Ay-GAL4, UAS-GFP/Ay-GAL4, UAS-GFP; +/+

  9. UAS-bskDN/UAS bskDN; P{y[+t7.7] v[+t1.8]=TOE.GS00536}attP40/CyO; UAS-dCas9/UAS-dCas9

All Stocks and crosses were maintained at 25°C, except for those used in RNAi based and GAL4-UAS expression experiments. In those cases, crosses were maintained at 29°C. For GAL80ts experiments, crosses were initially maintained at 18°C for 5 days (equivalent to 60 hr at 25°C), and then shifted to 29°C till dissection (Figure 2—figure supplement 1D).

For synchronization of larvae, flies were allowed to lay eggs for 2 hr and newly hatched larvae within 1 hr interval were collected and transferred onto fresh food plates and aged for specified time periods at 25°C.

Metabolic supplements and inhibitors

Request a detailed protocol

Fatty acid β-oxidation inhibitors: Etomoxir (Cayman Chemicals, Cay11969, inhibitor of CPT1) and Mildronate (Cayman Chemicals, Cay15997, inhibitor of carnitine biosynthesis and transport) were used at a concentration of 5 µM and 100 µM respectively mixed in fly food and fed to larvae from 48 hr AEH and analysis of lymph gland was done in late third instar stages. L-carnitine hydrochloride (Sigma-Aldrich, C0283) at a concentration of 100 mM has been used to augment FAO by allowing the entry of palmitic acid into the mitochondria. L-carnitine was used at 100 mM concentrations in fly food and fed to larvae for 48 hr in third instar analysis and for 24 hr in second instar analysis. Glycolytic inhibitor: 2-DG (2-Deoxy-D-Glucose (2-Deoxyglucose) (Sigma-Aldrich, D8375) used at a concentration of 100 mM mixed in fly food and fed to larvae for 48 hr in third instar analysis and for 24 hr in second instar analysis. Sodium acetate (Sigma-Aldrich, 71196) supplement was used at a concentration of 50 mM and fed to larvae from second instar 36 hr AEH onwards and analysis was done in late third instar stages. Similar aged larvae fed on vehicle controls served as control larvae. For all feeding experiments control larvae had same vehicle control level mixed in fly food. Fly food mixed with permissible food dye was fed to the control and experimental larvae and larvae with abundant food intake were picked for the experimental analysis.

Clonal analysis using flp-out clone using Ay-GAL4 system

Request a detailed protocol

Generation of clones was done by the Ay-GAL4 system that combines the technique of Flippase (Flp)/FRT system and the GAL4/UAS system (Ito et al., 1997). In this system, the Act5C promoter GAL4 fusion gene is interrupted by a FRT cassette containing yellow (y+) gene. Heat shock treatment activates the Flp gene which in-turn excises the FRT cassette between the Act5C promoter and GAL4 sequence. This activates the expression of Act5C-GAL4 in cells. To induce UAS-whd RNAi and UAS-bskDN clones, mid second instar larvae of genotypes: hsFlp; Ay-GAL4, UAS-GFP; UAS-whd RNAi and hsFlp/UAS-bskDN; Ay-GAL4, UAS-GFP were subjected to heat shock for 90 min at 37°C, respectively. Post heat shock, larvae were transferred to 25°C to recover for 2 hr, then to express the respective knockdown constructs, larvae were reared at 29°C till dissection.

Immunohistochemistry and imaging

Request a detailed protocol

The primary antibodies used in this study includes mouse anti-P1 (Kurucz et al., 2007), rabbit anti-Pxn (J. Fessler), rabbit anti-proPO (M. Kanost), rat anti-DE Cadherin (Cat# DE-cad (DE-cadherin), RRID:AB_2314298, 1:50, DSHB), rat anti-Ci155 (Cat# 2A1, RRID:AB_2109711, 1:2, DSHB), rabbit anti-GFP (Cat# A-11122, RRID:AB_221569, 1:100, Invitrogen), rabbit anti-H3 (Cat# 9927, RRID:AB_330200, 1:400, Cell Signaling Technologies), rabbit anti-H3K9 acetylation (Cat# 9927, RRID:AB_330200, 1:300, Cell Signaling Technologies), rabbit anti-H4 pan acetylation (Cat# 06–598, RRID:AB_2295074, 1:500, Merck-Millipore). The following secondary antibodies mouse Cy3 (Cat# 115-165-166, RRID:AB_2338692), mouse FITC (Cat# 715-096-151, RRID:AB_2340796), rabbit Cy3 (Cat# 711-165-152, RRID:AB_2307443), rabbit FITC (Cat# 111-095-003, RRID:AB_2337972), rat Cy3 (Cat# 712-165-153, RRID:AB_2340667) from Jackson Immuno-research Laboratories were used at 1:400.

Lymph gland from synchronized larvae of required developmental age was dissected in cold PBS (1X Phosphate Buffer Saline, pH-7.2) and fixed in 4% Paraformaldehyde (PFA) for 45 min (Mandal et al., 2007) at room temperature (RT) on a shaker. Tissues were then permeabilized by 0.3% PBT (0.3% triton-X in 1X PBS) for 45 min (3 × 15 min washes) at RT. Blocking was then done in 10% NGS, for 30–45 min at RT. Tissues were next incubated in the respective primary antibody with appropriate dilution in 10% NGS overnight at 4°C. Post incubation in primary antibody, tissues were washed thrice in 0.3% PBT for 15 min each. This was followed by incubation of tissues in secondary antibody overnight at 4°C. The tissues were then subjected to four washes in 0.3% PBT for 15 min each, followed by incubation in DAPI solution (Invitrogen) for 1 hr at RT. Excess DAPI was washed off from the tissues by 1X PBS before mounting in Vectashield (Vector Laboratories).

Immunohistochemical analysis of histone acetylation in lymph gland

Request a detailed protocol

Immunostaining for specific histone acetylations were performed with a slight modification of the above protocol. Lymph gland from synchronized larvae was dissected in ice cold PBS with deacetylate inhibitors (Sodium butyrate (10 mM, EMD Millipore, 19–137) and Nicotinamide (10 mM, Sigma-Aldrich, 72345) and fixed in 4% PFA prepared in ice-cold 1X PBS (pH 7.2) for 5 hr at 4°C. Tissue were then permeabilized by 0.3% PBT for 45 min. Blocking was done with 5% BSA made in 1X PBS. Primary antibody and secondary antibody incubation solutions were made in 5% BSA in 1X PBS and subsequent washings were done with 0.1% PBT.

Immunohistochemical analysis of DE-cadherin expression in lymph gland

Request a detailed protocol

To detect the DE-cadherin expression, lymph glands were incubated in DE-Cadherin antibody (1:50 in PBS) before fixation (Langevin et al., 2005) for 1 hr at 4°C. Tissues were then fixed in 4% PFA prepared in ice cold 1X PBS (pH 7.2) for 5 hr at 4°C. Then, tissues were washed thrice with 0.3% PBT for 30 min. Secondary antibody incubation, washes, and mounting were performed following the standard protocol (Sharma et al., 2019).

Streptavidin-Cy3 labeling of mitochondria

Request a detailed protocol

Larvae were dissected in cold PBS followed by fixation in 4% PFA overnight at 4°C. This was followed by permeabilization with 0.1% PBT (0.1% triton-X in 1x PBS) for 45 min at RT and incubation in Streptavidin-Cy3 in 1:200 dilution (Molecular Probes, 434315) in 1XPBS for 1 hr at RT in dark. Post incubation samples were, washed thrice in PBS for 30 min. Lymph glands were then mounted in Vectashield and imaged in Zeiss LSM 780 confocal microscope.

EdU labeling

Request a detailed protocol

Click-iT EdU plus (5- ethynyl-2’- deoxyuridine, a thymidine analog) kit (Invitrogen, C10639) plus was used to perform DNA replication assay. Lymph glands were dissected and incubated in EdU solution (1:1000 in PBS) for 40 min at RT for EdU incorporation. Next fixation was done in 4% PFA prepared in 1X PBS (pH 7.2) for 45 min at RT. Tissue were then permeabilized by 0.3% PBT (0.3% triton-X in 1X PBS) for 45 min at RT. Blocking was then done in 10% NGS, for 30–45 min at RT. To detect the incorporated EdU in cells, azide-based fluorophore were used as described in manufacturer protocol. EdU-labeled cell counting was done using spot detection function in Imaris Software.

2-NBDG assay

Request a detailed protocol

The protocol was slightly modified after (Zou et al., 2005). Larvae were dissected in ice-cold PBS and incubated in PBS with 0.25 mM 2-NBDG (Invitrogen, N13195) for 45 min at RT, washed twice in PBS for 5 min, fixed 45 min in 4% PFA and washed twice for 10 min in PBS. All washes and the fixation were done with ice-cold PBS (4°C). Lymph glands were speedily dissected and mounted in Vectashield and were imaged immediately with a Zeiss LSM 780 confocal microscope.

Lactate dehydrogenase assay

Request a detailed protocol

Lactate dehydrogenase in vivo staining was modified from Abu-Shumays and Fristrom, 1997. Lymph glands of wandering third instar larvae were dissected in cold 1X PBS (pH8). Samples were fixed for 25 min in 0.5% glutaraldehyde in 1X PBS at room temperature, followed by four washes in 1X PBS for 15 min each. Staining was performed at 37°C in a solution of 0.1M NaPO4 (pH 7.4), 0.5 mM lithium lactate, 2.75 mM NAD+, 0.5 mg/ml NBT (Nitro blue tetrazolium) with 0.025 mg/ml PMS (Phenazine methosulfate). Reaction was stopped by washing in cold 1X PBS having pH 7.5. The samples were washed in four 1X PBS washes of 5 min each and immediately mounted and imaged.

LipidTOX staining

Request a detailed protocol

Larvae were dissected in cold PBS followed by fixation in 4% PFA for 1 hr at RT, permeabilized by 0.1% PBT (0.1% triton-X in 1X PBS) for 45 min at RT. It was then incubated in 1X LipidTOX (diluted from 1000X stock provided by the manufacturer; Molecular Probes, H34477) in PBS for 1 hr at RT in dark, washed thrice in PBS for 30 min. Lymph glands were then mounted in Vectashield and imaged in Leica SP8 confocal microscope.

Nile red staining

Request a detailed protocol

Larvae were dissected in cold PBS followed by fixation in 4% PFA for 1 hr at RT, permeabilized by 0.1% PBT (0.1% triton-X in 1X PBS) for 45 min at RT and incubated in 0.5 ug/mL Nile red (Molecular Probes, N1142) in PBS for 1 hr at RT in dark, washed thrice in PBS for 30 min. Lymph glands were mounted in Vectashield and imaged in Leica SP8 confocal microscope.

Detection of ROS

Request a detailed protocol

Larvae were dissected in Schneider’s medium (Gibco, 21720001) followed by incubation in 0.3 µM DHE (Molecular Probes, D11347) in Schneider’s medium for 8 min at room temperature in dark. This was followed by two washes in 1X PBS for 5 min each; a brief fixation was done with 4% PFA for 10 min followed by two quick 1X PBS washes. Tissues were then mounted in Vectashield and imaged in Zeiss LSM 780 confocal microscope.

Imaging and statistical analyses

Request a detailed protocol

Images were captured as confocal Z-stacks in Zeiss LSM 780, Leica SP8 confocal, and Olympus Flouview FV10i microscopes. Same confocal imaging settings were employed for image acquisition of control and experimental samples related to an experiment. Each experiment was repeated with appropriate controls at least three times to ensure reproducibility of the results. Data expressed as mean+/-Standard Deviation (SD) of values from three sets of independent experiments in GraphPad. Each dot in GraphPad represents a data point. Graphs plotted in EXCEL have Error Bars representing the Standard Deviation while graphs plotted in GraphPad employs Error Bars as mean+/-Standard Deviation. At least 10 images were analyzed per genotype, and statistical analyses performed employed two-tailed Student’s t-test. Raw data related to statistical analysis are attached in the source file of each figure along with graphs plotted in excel.

p-Value of <0.05;<0.01 and<0.001, mentioned as *, **, *** respectively are considered as statistically significant while n.s. = not significant.

Quantitative analysis of differentiation index in lymph gland

Request a detailed protocol

To measure the differentiation index of the primary lobe of lymph glands, middle confocal Z-stacks of a lymph gland image covering the Medullary Zone (MZ) were merged into a single section using ImageJ/Fiji (NIH) software as previously described (Shim et al., 2012). The merged section reflects the differentiated cell and hemocyte progenitor area clearly. For images with more than one fluorophore channel, each channel was separately analyzed. To measure differentiation index, P1-positive area was recalibrated into an identical threshold by using the Binary tool (Process–Binary–Make binary, Image J). Wand tool was used to capture the area with identical threshold whereas the size was measured using the Measure tool (Analyse–Measure). To measure the total area of one primary lobe of lymph gland, recalibration of the total area was then done by the Threshold tool until it was overlaid with identical threshold colour. Wand tool was used for selecting the total area for measurement. Differentiation index/fraction was estimated by dividing the size of the P1/Pxn positive area by the total size of the primary lobe. At least 10 lymph glands were analyzed per genotype, and two-tailed Student’s t-test was done to evaluate the statistical significance.

Quantification of the number of EdU+, FUCCI+ and progenitor subpopulations

Request a detailed protocol

Counting the number of EdU+ and FUCCI+ progenitors in lymph glands was done as described earlier (Sharma et al., 2019), using spot detection and surface tool in Imaris software and normalized by total number of nuclei per primary lobe of lymph gland. Different progenitor sub-populations in lymph glands were counted using the surface and spot detection function in Imaris as illustrated in detail (Sharma et al., 2019) and normalized by total number of cells (nuclei) in primary lobe and represented as percentage of progenitors in each primary lobe. Using surface tool, surface is created over Dome+ progenitors and nuclear label channel is masked in Dome+ surface. By utilizing the spot detection tool, the number of nuclei is counted in Dome+ surface. Similarly, number of nuclei (DAPI/Hoechst) is counted in another surface created over Pxn+ cells. Next, in Dome+ surface, Pxn+ channel is masked. A surface is created over Dome+ Pxn+ region and nuclear channel is masked in this surface. Using spot detection tool, number of Dome+ Pxn+ IP nuclei are counted and its percentage can be calculated from the total number of nuclei in the primary lobe of lymph gland.

ATP assay

Request a detailed protocol

ATP assay was performed with three biological replicates from late third instar larvae. Whole larvae were homogenized in ATP assay Lysis buffer (Costa et al., 2013). The samples were boiled at 95°C for 5 min and diluted 1:100 in dilution buffer provided in ATP bioluminescence kit HSII (Sigma, 11699709001). Further assay was performed as per manufacturer protocol in Glomax 96 microwell Luminometer (Promega). Standard curve was generated and ATP concentrations were calculated. The ATP concentration was normalized with protein concentration and expressed in percentage to plot the graph in EXCEL.

Histone extraction and detection by western blotting

Request a detailed protocol

Histone from late third instar larvae were extracted using Histone extraction kit (Abcam, ab113476) following manufacturer protocol and quantitated by Bradford reagent (Biorad, 5000006). Equal amount of protein of each genotype was run on 4–12% SDS-PAGE and transferred to PVDF membrane (Millipore, IPVH00010). Blots were developed using Luminata Crescendo Western HRP substrate (Millipore, WBLUR0500) in LAS2000 blot imaging instrument. Primary antibodies used rabbit anti-H3 (Cat# 9927, RRID:AB_330200, 1:1000, Cell Signaling Technologies), rabbit anti-H3K9 acetylation (Cat# 9927, RRID:AB_330200, 1:1000, Cell Signaling Technologies). Secondary antibody rabbit anti-IgG-HRP (Cat# A00098-1 mg, RRID:AB_1968815, 1:5000, GenScript) was used. The band intensity was measured in Image J and normalized with histone H3 as loading controls. Analysis was done using three biological replicates.

Quantitative RT-PCR

Request a detailed protocol

Extraction of RNA from lymph gland was performed from late third instar larvae of each genotype using TRIzol (Invitrogen, 15596018) followed by RNAeasy Mini Kit (Qiagen, 74104) according to the manufacturer’s instructions. cDNA was prepared using the Verso cDNA Synthesis Kit (Thermo Scientific, AB1453B). To quantitate transcripts, qPCR was done using iTaq Universal SYBR Green Supermix (Biorad, 1725124) on a CFX96 Real-Time system/C1000 thermal Cycler (Biorad). Drosophila Actin5C was used as internal control.

Analysis was done using at least three biological replicates.

References

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10
  11. 11
  12. 12
  13. 13
  14. 14
  15. 15
  16. 16
  17. 17
  18. 18
  19. 19
  20. 20
  21. 21
  22. 22
  23. 23
  24. 24
  25. 25
  26. 26
  27. 27
  28. 28
  29. 29
  30. 30
  31. 31
    The Drosophila mushroom body is a quadruple structure of clonal units each of which contains a virtually identical set of neurones and glial cells
    1. K Ito
    2. W Awano
    3. K Suzuki
    4. Y Hiromi
    5. D Yamamoto
    (1997)
    Development 124:761–771.
  32. 32
  33. 33
  34. 34
  35. 35
  36. 36
  37. 37
  38. 38
  39. 39
  40. 40
  41. 41
  42. 42
  43. 43
  44. 44
  45. 45
  46. 46
  47. 47
  48. 48
  49. 49
  50. 50
  51. 51
  52. 52
  53. 53
  54. 54
  55. 55
  56. 56
  57. 57
  58. 58
  59. 59
  60. 60
  61. 61
  62. 62
  63. 63
  64. 64
  65. 65
  66. 66
  67. 67
  68. 68
  69. 69
  70. 70
  71. 71
  72. 72
  73. 73
  74. 74
  75. 75
  76. 76
  77. 77
  78. 78
  79. 79
  80. 80
  81. 81
  82. 82
  83. 83
  84. 84
  85. 85
  86. 86
  87. 87
  88. 88
  89. 89
  90. 90
  91. 91

Decision letter

  1. K VijayRaghavan
    Senior and Reviewing Editor; National Centre for Biological Sciences, Tata Institute of Fundamental Research, India
  2. Yukiko M Yamashita
    Reviewer; University of Michigan, United States

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

Acceptance summary:

In this manuscript, Mandal and colleagues explore the function of FAO (fatty acid oxidation) in differentiation of hematopoietic progenitors in Drosophila larval hematopoietic organ, lymph gland (LG). They show that supplementation with acetate rescues the phenotype associated with defective FAO, and that FAO is downstream of JNK, a known regulator of differentiation, providing insights into how metabolic pathway may be intertwined with differentiation pathway.

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 "Fatty acid β-oxidation regulates hemocyte progenitor homeostasis in Drosophila larval lymph gland" for consideration by eLife. Your article has been reviewed by three peer reviewers, and the evaluation has been overseen by a Reviewing Editor and a Senior Editor. The following individual involved in review of your submission has agreed to reveal their identity: Yukiko M Yamashita (Reviewer #1).

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 for publication in eLife. Should they be able to address all substantive comments and choose to consider eLife, a fresh submission can be examined, without any guarantees.

Our consultation made every effort to understand the important assertions made and whether they were borne by the data presented. While the reviewers started with different perspectives, we have converged to a common position. We suggest that the authors go through our preamble, which captures this converged view, and also the specific concerns of each reviewer. While some of our major concerns can be addressed by substantive re-writing, cutting down generalisations and over-interpretation others require experimental attention.

One important issue, which was reiterated in our consultations, is exemplified by the following statement and related data "We show that the self-renewing hemocyte progenitors prefer glycolysis while the quiescent progenitors adopt Fatty acid β-oxidation (FAO) for their differentiation. Perturbation of β-oxidation in progenitors results in loss of quiescence…." Though the authors' primary question is centered on distinguishing metabolic states within progenitors, the authors do not identify these. Generic results are interpreted as applying to specific populations and extended to homeostasis.

Staging larvae and using appropriate established Gal4 drivers is, of course, acceptable, but the authors have not done this properly. Unlike, for example, the wing and eye discs, the larval LG do not have a uniform, reproducible structure or pattern. One cannot delineate zones morphologically or by using a single marker. Progenitor sub-populations (dome negative pre-progenitors, dome+ progenitors and dome+Pxn+hml+ intermediate progenitors) (reviewed in Banerjee et al., Flybook2019) and the three types of differentiated cells (plasmatocytes, crystal cells and lamellocytes) are functionally and metabolically different. However, the authors do not distinguish between these populations in their analysis. For example, mitochondrial analysis is shown at low resolution and in an undefined population (Figure 1L, M) and hml>GFP is used only as a marker of differentiation (Figure 4).

The assertion of the manuscript is that they are distinguishing between progenitor types, which they call self-renewing and quiescent. This can be done only upon identifying and characterising subpopulations. Further, since they incorrectly use the presence or absence of proliferation marks as a readout for self-renewal and quiescence, it is not clear how the claim that these are two metabolically distinct precursor states is valid. Rather, based on the analysis done (EdU, Figure 1B,C) they are simply proliferating or non-proliferating cells (which may or may not be progenitors), the latter poised for, or in the process of differentiation. Expectedly, their metabolic states differ- and the authors show that these are glycolysis for the former and FAO for the latter. Figure 1B,C: TepIV and EdU labelling – it looks as if there are co-labeled cells, but these are ignored. Also TepIVGAL4 reporter does not label the entire progenitor pool. The existence of self-renewing or quiescent progenitors has not been formally demonstrated in the LG- if the authors had demonstrated their existence, this would have been very interesting indeed.

The metabolic fluctuations we are concerned about are intra-LG. This is a major challenge in characterizing metabolic states of heterogeneous progenitors. Since all analysis is done only with one progenitor marker at a time (mainly dome) which is also expressed in early stages of differentiation, or with no marker at all, (Example Figure 1 ; Figure 1—figure supplement 1, compare panels A and C) claims regarding metabolic differences are not substantiated. As is also evident from the images in Figure 1, the same tissue from the same genotype and stage varies greatly in terms of size, shape and organisation, hence rigorous quantitation with additional markers is essential.

The authors seem confused about what the focus of their study is. They state in the Abstract that the major outcome of the study is coupling FAO with epigenetic modifications linking to regulation of differentiation by JNK- while this part of the study has merit, it is not very relevant to the primary question they pose about blood cell progenitor homeostasis.

The reviews give a very detailed analysis, including comments for almost every panel of Figure 1. However, some examples are expanded on a couple of specific cases below.

Figure 1D, E states that self-renewing early stage progenitors exhibit elevated levels of glucose uptake compared to late-stage quiescent progenitors. In D there is no marker for progenitors or differentiated cells- clearly there are varying intensities of signal. How can one assume that all 2-NBDG cells are progenitors, early or late? Further, given variation between preps, how does one compare levels between second and third instar without having a control to normalise it or assess relative levels between progenitors and differentiated cells? Assuming second and third instar preps were stained and imaged in the same experiment, how much variation is seen? A graph representing this is essential as the lymph gland morphology can vary greatly depending on the status of the larval cultures. Additionally, a proliferation assay is required along with the uptake or demarcation of the MZ/CZ boundaries to show reduced uptake in the pro-hemocyte pool of the 3rd instar lymph gland.

Figure 1F, G – To say that there is high Glut-1 GFP in proliferating progenitors of the second instar, labelling with a proliferation marker, progenitor marker, and differentiation marker is required along with quantitation to show that High Glut 1-GFP is expressed in proliferating progenitors.

Figure 3E-G, the authors claim a higher number of EdU incorporation is seen in the progenitors, however, the images provided show higher EdU signal at the periphery; also CZ/MZ boundary is not demarcated, and progenitor markers are not used.

If the results about FAO and its downstream effects in the LG are de-linked from progenitor sub-types and homeostasis, this would be more acceptable. Currently, two parts of the manuscript (progenitor types and metabolic status connecting to signaling) are not well connected. The authors may want to consider that they are perhaps best dealt separately and not in the same manuscript.

Reviewer #1:

This study by Mandal and colleagues studied the role of metabolic circuitry in stem cell maintenance/differentiation. General impacts of metabolism on cell fates, particularly those between stem cells and their progenies, have increasingly drawn researchers' attention in recent years. By using Drosophila hematopoietic progenitors (in lymph gland), the authors explored the role of metabolic circuitry in this lineage. The main conclusion is that actively proliferating stem cells utilize glycolytic pathway whereas quiescent stem cells favor fatty acid β-oxidation, which promotes differentiation.

Furthermore, they propose that FAO generates acetyl-CoA, which not only generates energy through TCA cycle, but can contribute to protein (e.g. histone) acetylation. Perturbation of FAO leads to defects in quiescence (progenitor starts proliferating and cannot differentiate). Upregulation of FAO leads to stem cell quiescence and precocious differentiation. The authors link this observation to histone acetylation and thus epigenetic regulation.

Figure 1: Does blocking glycolysis have any impact on differentiation? Does it only affect progenitor proliferation without impacting later differentiation?

Figure 2: blocking FAO by means of whd knockdown resulted in increased progenitors and decreased differentiating cells. Further in Figure 3, they show that loss of FAO also increases progenitor proliferation and thus progenitors are not entering quiescence. Collectively these results indicate that FAO is required for progenitors to enter quiescence, and thus their “differentiation defect” is secondary to their inability to enter quiescence. These results are quite clear, but when I was reading the description for Figure 2, I was misled to think that FAO blocks differentiation. Any ways to make this transition (Figure 2 to Figure 3) a bit less misleading?

In Figure 4, they show that upregulation of FAO is sufficient to induce quiescence and subsequent differentiation in progenitors.

Figure 5 demonstrate the inter-relationship between glycolysis and FAO. Defective FAO led to sustained glycolysis.

Figure 6-7 slightly shift in the focus and now addresses FAO's function in histone acetylation. The authors now suggest that it is the key to differentiation.

Figure 8 talks about JNK pathway in progenitor differentiation.

Although the authors' claims are well supported in general, Figures 6-8 make me wonder whether FAO contributes to progenitor cell differentiation through metabolic aspects at all. Although each piece of discoveries in this paper is important, I don't know if the authors can claim that FAO regulates progenitor differentiation through BOTH metabolic regulation and histone acetylation. Especially because Figure 7 shows that acetate supplementation is enough to rescue whd mutant phenotype, one has to wonder whether “metabolic” aspect ever contributes to self-renewal/differentiation choices. And if this is all about histone acetylation, what is the importance of glycolysis in self-renewing progenitors? Of course, it is an intriguing possibility that this is nature's elegant solution, natural shifts in metabolism also triggers epigenetic programming. I am not asking to do experiments to tell apart the possibilities, but the fact that their data do not necessarily show the requirement of FAO regulating metabolism per se in regulating differentiation, and the possibilities have to be carefully discussed.

I understand it might be technically difficult, but can they tell that histones are less acetylated in self-renewing progenitors compared to quiescent progenitors?

I think that Figure 8 is unnecessary. This does not add to the major message of the paper.

Reviewer #2:

Tiwari et al. present data that they claim suggests a potential role for FAO in metabolic regulation of progenitors in Drosophila larval lymph gland. They interpret their data to say that self-renewing and quiescent progenitors in the LG differ in their metabolic state. The latter exhibit FAO which results in acetylCoA production and thereby increase in histone acetylation and altered gene expression. Pharmacological increase in acetylation has the same effect whereas reduced histone acetylation maintains progenitor self-renewal. They state that the major outcome of the study is coupling FAO with epigenetic modifications linking to regulation of differentiation by JNK.

The metabolic state of blood cell progenitors has been studied extensively in vertebrates and to a limited extent in Drosophila. This study further characterizes progenitor sub-populations for their metabolic status. However, a major shortcoming of the manuscript is the misinterpretation of literature on which their hypothesis and analyses are based. As a result, the fundamental question being addressed about self-renewal and quiescence in LG progenitors is flawed. Further, there are systemic problems with the methods used to analyze, present and quantify data in this manuscript that make it impossible for me to judge the validity of their conclusions. Controls are rarely presented and cell sub-populations are not identified or marked. It is standard practice to include a control for all gene expression especially when reported for the first time and also to identify and distinguish between cell sub-populations when differences in their properties are claimed. Furthermore, the data in this paper are difficult to interpret and I am confused by how the authors came to their conclusions.

There is a general lack of quantification that would be required to make the interpretations presented.

Hence the first part (Figure 1) that emphasizes self-renewal and quiescence distracts from the main findings and the data are unclear and analyses questionable. Since the role of JNK in LG differentiation and its connection to metabolism are already established, the results presented here are only incremental to our understanding of LG progenitor maintenance. Further, a lot of the experiments related to this section will have to be repeated with appropriate markers and controls and quantitated rigorously to be convincing. Hence I think the present manuscript is not suitable for eLife.

A detailed review is provided below:

Introduction:

The authors seem to be unaware that of the different hematopoietic populations in vertebrates and how they compare with that of Drosophila. A major problem is the idea that there are self-renewing and quiescent progenitors in the larval LG. It is important to keep in mind that the LG is a powerful but limited model of vertebrate hematopoiesis. Hence a one-on-one correlation between cell types, states and metabolic status should not be made. The Abstract should be corrected to avoid the use of terms such as self-renewal and quiescence. The authors start the Introduction with the aim of testing the relation between metabolic intermediates and histone modifications but deviate to self-renewal and quiescence.

Introduction section

Studies on various stem cell scenarios primarily in Hematopoietic Stem Cells(HSCs) have established that various states adopted by stem cells like, quiescence, proliferation, and differentiation are liable to different metabolic requirements…

This suggests that a stem cell can occur in multiple states and is dynamically switching between them. However, this is not the case. Note that the references cited are for multiple sub-populations of vertebrate HSC. There has been no formal demonstration of stem cells in the LG nor of self-renewal. However, the authors interpret all their data in the context of self-renewal and quiescence and hence this is incorrect and misleading. Further, the experimental evidence does not support the conclusions (see below).

Introduction paragraph two

The blood progenitors found in Drosophila late larval lymph gland are akin to the vertebrate common myeloid progenitors (CMP)(Owusu-Ansah and Banerjee, 2009). They are quiescent, have high levels of ROS (Owusu-Ansah and Banerjee, 2009), lack differentiation markers, and can give rise to all Drosophila blood lineages.

This again is a misinterpretation. Firstly, CMPs are not quiescent, only long-term repopulating HSCs are proven to be quiescent. LG progenitor quiescence has not been conclusively demonstrated either. The authors interpret lack of proliferation as quiescence, these are two very different cellular states and associated with different gene expression profiles as well as cell phenotype. Finally, unlike in vertebrate studies, the ability of a LG progenitor to give rise to "all Drosophila blood lineages" has not been demonstrated.

Introduction paragraph three:

This should be edited to indicate that not all transcription factors are conserved between the two systems.

The authors equate proliferation with self-renewal- these are two different processes that share gene networks. EDU labelling indicates proliferation not self-renewal. Self-renewal is a very special property of stem cells that can also differentiate. Similarly, absence of EDU label indicates the cells are not proliferating or are cycling slowly- quiescence requires existence in the G0 phase, which has not been demonstrated in the LG.

Introduction paragraph four:

Dynamic states of self-renewal, quiescence and differentiation.

Note that a single cell does not switch between these phenotypic states. Differentiation is normally a one-way street down the lineage through each state, and necessitates metabolic shifts.

The data do not provide evidence that the progenitors primed to differentiate rely on FAO, this is merely correlative.

General comments on results:

Pharmacological treatments: This is another major point that the rest of the paper relies on, yet the authors again show no controls and inadequate quantification. While the entire paper rests on analysis of a few hundred cells in the LG, the Western blot analysis is done with the whole larva and results are interpreted in the context of the LG.

The authors claim that there are metabolic differences between progenitor sub-populations that correlate with potential to differentiate. However pharmacological treatments such as acetate feeding is done at the level of the whole animal. How much of acetate is sensed by the progenitors in question? how does a general increase in acetylation affect histone in various tissues including the LG? As systemic signals are key to maintaining blood progenitor homeostasis, the effects seen could be due to cell extrinsic factors. The conclusion that this treatment proves the role of FAO and histone acetylation in progenitor quiescence is far-fetched. Further there is no evidence of that glycolytic and FAO metabolizing sub-population are differentially affected.

It is similarly unclear in which progenitor sub-populations whd and JNK signaling are affected. Quantification is required for differentiation checked by P1 and progenitor status should be checked by dome; also differentiation to crystal cells and lamellocytes should be examined. This is the first report of metabolic requirement for progenitor maintenance; a thorough analysis of various differentiation markers is essential.

Figure wise comments

Figure 1 is fraught with misinterpretations. (Results paragraph one and two). Further there is no quantitation whatsoever and a lack of controls. Two different developmental stages are compared and cell identity is arbitrarily assigned. No markers are used for self-renewing or quiescent cells and it is not clear how the authors identify these. Edu positive cells are only a subset of the second instar LG (these could be differentiating) and seen only in the CZ of the third instar.

Figure 1D, E states that self-renewing early stage progenitors exhibit elevated levels of glucose uptake compared to late stage quiescent progenitors. In this case a proliferation assay is required along with markers to demarcate the MZ/CZ boundaries to show that the uptake is less in the prohemocyte pool of the 3rd instar lymph gland. It is likely that the 2NBDG labeled cells could be late progenitors in the second instar that persist in CZ. Without appropriate markers this cannot be distinguished. Further third instar progenitors may be slower proliferating or arrested in the cell cycle- this is not quiescence. A cell cycle analysis needs to be done through L2 and L3 to resolve the two states, so called self-renewal and differentiation.

What the authors call quiescence is likely a non-proliferative state that precedes differentiation, which is well documented in many developmental contexts. Similarly, as mentioned in Results paragraph two, glycolysis is seen in rapidly proliferating cells, such as cancer cells. This is not necessarily self-renewal.

GlutGFP expression is patchy (Figure 1F, G) and not clear whether it is in or around the cells, this looks more like background. A no primary antibody control and a Glut RNAi control are required. MZ should be marked by dome or DEcad and perform EdU labeling in order to show that High Glut 1-GFP is expressed in proliferating progenitors. Similarly, aldolase expression (supplement 1A, B) is not convincing and is seen only in the periphery of the second instar.

Figure 1 H, I imaging plane is unclear. It is not clear what the authors are trying to show. Imaging parameters are not uniform.

Figure 1 K: Which area of the LG is shown? Nuclear staining should be included.

Figure 1 L-O', it is impossible to comment on the status of mitochondrial network with this analysis. Which part of the LG is shown? High resolution images with live tracking and video microscopy are required to analyze mitochondrial morphology, length and dynamics. Without detailed information regarding these parameters in L2 and L3 progenitors and in differentiated CZ cells, one cannot make any comparisons as there is a large variation in mitochondrial size and morphology within any population. This needs co-analysis with progenitor and differentiation markers and thorough quantitation- CZ/ MZ mitochondrial pattern needs to be mapped first.

There is no data to show mitobiogenesis.

Figure 1Q, R: the accumulation/increase of Lsd-2 in the 3rd instar lymph gland would be clear if 2nd instar images are presented for Nile Red.

Figure 1 requires quantification and insets indicating region of interest; double labeling with progenitor marker and or CZ markers or EdU labeling is essential.

Figure 2: Dapi staining or demarcation of LG boundary as well as MZ/CZ marker staining should be shown. Crystal cell and lamellocytes differentiation status should also be analyzed.

2H,L Q Graph is misleading, it suggests there is a decrease in dome+ area. Y axis should indicate plamatocyte % or area. What is differentiation index?

Figure 3E-G: the authors claim higher number of EdU incorporation is seen in the progenitors, however the images provided show higher EdU signal at the periphery; also CZ/MZ boundary is not demarcated, progenitor marker should be used and data quantitated.

Figure 4F, G quantitation required for commenting on precocious differentiation (Hml+ve cells).

Figure 5F-H quantitation required for EdU labeling to show that feeding with 2-DG rescues the hyperproliferation phenotype.

Figure 6A-C claims loss of Histone Acetyl Transferase (HAT), Gcn5 and chm phenocopies differentiation defect seen in the progenitors of FAO loss of function. What is the effect on proliferation in panels 6A-G. Does it corroborate with active proliferation seen in whd mutants.

For Figure 7F materials method suggests western blots were done from larva of the respective genotypes, this cannot be correlated to the LG effects. It would be more appropriate to show H3K acetylation in situ in the lymph gland.

Figure 7 H-I Lymph gland images required in order to visualize that the addition of H3K acetylation increases in L-carnitine fed v/s unfed larvae.

Figure 8A- C, expression of dominant negative allele of basket (bsk, Drosophila ortholog of Jun-Kinase) results in a compromised differentiation analogous to whd knockdown. Can the phenotype be rescued by feeding L-carnitine or acetate?

The effect of Jun (bsk) on whd is shown at the transcriptional level. Does Jun (bsk) affect enzymes of the FAO pathway? Also Jnk is known to activate Foxo, what is the status of Foxo reporter (Thor-LacZ) in whd mutants.

Foxo overexpression advances differentiation (Owusu-Ansah and Banerjee, 2009), does overexpression of Foxo affect/ rescue whd mutant phenotype by inducing differentiation.

Reviewer #3:

Mandal and colleagues use the lymph gland of Drosophila to identify the genetic events by which a metabolic switch (from glycolysis to fatty acid oxidation, FAO) drives a developmental transition (from progenitor to differentiated cells). They first combine fluorescent glucose uptake analysis, expression of metabolic enzymes and glucose transporters and exposure to a glucose antagonist to present evidence that early but not late lymph glands make use of glucose to grow. They next combine genetic analysis and pharmacological treatments to present functional evidence that a metabolic switch to FAO is required in late lymph gland cells to differentiate. FAO requirement does correlate with a change in the protein levels of Ci but appears to act downstream of the differentiation signal ROS. Genetic or pharmacological premature induction of FAO in early lymph glands is sufficient to drive progenitor cells into differentiation and the use of glucose is an absolute requirement for progenitor cells to keep proliferating in the absence of FAO. At the end of the manuscript, experimental data indicate that FAO is required in differentiated cells for histone acetylation downstream of ROS-induced JNK signaling.

The paper is well written (even though grammar should be improved), figures self-explanatory and the main message (from a metabolic to a developmental switch) and experimental setting (combination of genetics and pharmacology) is timely and appropriate for a journal like eLife.

[Editors’ note: further revisions were suggested prior to acceptance, as described below.]

Thank you for submitting your article "Fatty acid β-oxidation is required for the differentiation of larval hematopoietic progenitors in Drosophila" for consideration by eLife. Your article has been reviewed by three peer reviewers, and the evaluation has been overseen by K VijayRaghavan as the Senior and Reviewing Editor. The following individual involved in review of your submission has agreed to reveal their identity: Yukiko M Yamashita (Reviewer #1).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Summary:

Mandal and colleagues beautifully use the lymph gland of Drosophila to identify a role of FAO in the differentiation step from progenitor to differentiated blood cells. Authors combine a great variety of tools (drugs, reporters, inhibitors) together with powerful genetics (RNAi, CRISPR, mutants) to draw a line from JNK to Whd expression, from Whd to FAO, and from FAO to histone acetylation through the production of acetlyCoA. The present manuscript is a revised version of a previous one submitted to eLife and has improved in many aspects (experimental setting is very elegant, the combination of genetics and drugs/reporters is extraordinary and the epistatic analysis to demonstrate their claims is very robust). Others (writing, figure presentation, cartoons, etc) should be seriously improved in order to attract the attention of the general reader. While we strongly support publication of this manuscript in eLife, the following issues should be thoroughly addressed. Otherwise, even though the science is excellent, the quality of the manuscript will not be deserving to be published in eLife.

While the data are generally of good quality, there are significant gaps that need to be filled. The results must be seen in the context of what we already know – the link between JNK and FAO is known, as is the role of acetyl CoA in epigenetic modification. Tiwari et al. show that this also occurs in and is essential for blood progenitor differentiation. However, the same has been shown in vertebrate HSC. The authors conflate two very different cell types and approaches. Vertebrate HSCs have been extensively characterized and their quiescence, multipotency and repopulating ability has been demonstrated unambiguously. The same is not true for Drosophila hematopoiesis – unfortunately several studies try to force cellular parallels between the two. This is not required and Drosophila is an excellent hematopoietic model even though it does not have HSC.

Essential revisions:

Authors should talk to non-Drosophila colleagues to help them in thoroughly revising the following aspects of the manuscript:

1) Abstract is not well written and should be improved, specially the last two sentences. It does not summarize the major aspects of the work and will not attract the attention of any reader that has not a special interest in the Drosophila LG.

2) Cartoon depicted in Figure 1A and explanation in the Introduction do not have anything to do with each other (CZ only explained in the text and others only in the figure) in the figure and does not clarify anything. It's just confusing and should be completely changed and re-done.

3) Last paragraph of the Introduction should summarize the major findings and they do not so. References to the JNK/ROS axis are lacking. Jnk itself is not introduced. Again, this part of the manuscript is intended to attract the non-expert community

4) First part of the results: the why of using two developmental stages and different markers (pvf2, Dome, Hml, etc) is not clearly explained anywhere. Dome+, Dome+ Pxn+, Dome-, etc nomenclature is confusing (2N-O). Any way of improving it? Authors should be able to avoid technicalities.

5) Tens of mistakes (prolifeartes, progenitors experiences, etc) should be corrected.

6) Writing and conclusions in all sections should be improved.

Scientific issues:

7) Lipid markers and CG3902 expression. A Z section is required to demonstrate that the differential expression/incorporation levels between progenitor and differentiated cells is real not a consequence of the focal plane or thickness of the tissue.

8) Authors should use ubiquitous Gal4 drivers (NOT dome-Gal4) to demonstrate differences in the amount of mitochondria and cell cycle profiles (Fucci) between progenitor and differentiated cells.

9) Some negative results (changes in Hh and ROS levels w/o any impact on the process) should be moved to supplementary data.

10) CycB degradation of FlyFucci is not degraded in G1, as authors claim in subsection “Loss of FAO caused an increase in progenitor proliferation, and high redox levels”.

11) Single channels to monitor differences in EdU incorporation levels should be included (3G-I, 5M-P). Figure 5E, F is not described properly in the text.

12) Y axis in panel 3F: should be ROS and not Ci, right?

13) Materials and methods: Codes and/or references of all Fly stocks required. By the way, many of the fly lines used in the work are not included in Materials and methods. Stocks such as AyGal4 should be properly explained.

14) There is a lot of room for improvement in the organization of panels within Figures. Labels of most histograms are too small and error bars are not explained in the legends (SD or SDM?).

15) What is the trigger for differentiation? Is dietary lipid content a regulator of this as has been shown in vertebrates for regulation by FAO?

Mere carnitine supplementation is not sufficient as carnitine is not an essential nutrient, is abundant in the body under regular diet and carnitine deficiency results from mutations in its transport machinery. As the claim is that whd links JNK to FAO, and is transcriptionally regulated by JNK, a genetic rescue experiment with whd is essential, especially as Foxo overexpression increases differentiation. This manuscript often relies on compound treatment strategies alone despite having a genetically tractable model. Yet the role and effect of endogenously generated systemic cues is completely disregarded.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for resubmitting your work entitled "Fatty acid β-oxidation is required for the differentiation of larval hematopoietic progenitors in Drosophila" for further consideration by eLife. Your revised article has been evaluated by K VijayRaghavan (Senior Editor) and a Reviewing Editor.

The manuscript has been improved but there are some remaining issues that need to be addressed before acceptance, as outlined below:

Summary

In this revised manuscript, Mandal and colleagues explore the function of FAO (fatty acid oxidation) in differentiation of hematopoietic progenitors in Drosophila larval hematopoietic organ, lymph gland (LG). They show that supplementation with acetate rescues the phenotype associated with defective FAO, and that FAO is downstream of JNK, a known regulator of differentiation, providing insights into how metabolic pathway may be intertwined with differentiation pathway. They use abundant genetic models (mutants, reporters) to support their claim.

The revised manuscript by Tiwari et al. has addressed all the comments made on the previous version. The authors have repeated experiments, included controls, or done additional experiments where required. Importantly, progenitors have been correctly identified and analyzed. Also, progenitor-specific genetic analysis (perturbation of glucose uptake; downregulation of withered; upregulation of FAO in JNK depletion) has been done. The data are of good quality, presented clearly and gaps in analysis have been filled satisfactorily. The Discussion is much improved. This manuscript is significantly improved over the last submission and is an important contribution to the field.

However, there are still issues with the writing as following, which should be addressed prior to publication. Most of comments can be addressed by modifying writing, but some possible experiments are suggested : these experiments are just suggestions (in case they already have data) and not intended to be “essential” for revision, under the new editorial policy at eLife during the covid-19 pandemic.

Essential revisions

1) In the Introduction, the authors contrast hematopoietic stem cells (HSCs) in mammals with Drosophila hematopoietic progenitors, assuming that studying Drosophila hematopoietic progenitors will provide insights into how mammalian non-HSC progenitors are regulated by FAO. However, given that Drosophila does not have HSCs, and whether they represent HSCs or progenitors remain unclear (or whether it's comparison to mammalians HSC vs. progenitor is more relevant). In mammalian HSCs, FAO is required for stem cell maintenance, whereas in Drosophila LG, FAO is required for differentiation. I see that this difference may have driven the authors to make the above contrast, but I don't think it's a well-grounded argument and they should modify the Discussion accordingly.

2) Figure 1—figure supplement 1C shows reticular mitochondria as an indication of high FAO. Ideally, they should show less reticular mitochondria in other cell types in comparison.

3) Figure 4: to nail down the pathway, the epistasis analysis would be ideal, such as combination of genetic mutations (inhibit differentiation, such as whd mutant) and L-carnitine (drive differentiation). Especially, they conduct similar experiments in Figure 5, Figure 6, and Figure 8.

4) Figure 5I-L, 2-DG treatment rescues G2/M arrest defects in whd mutant, but does not seem to rescue differentiation defect. Any discussions?

https://doi.org/10.7554/eLife.53247.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.]

While the reviewers started with different perspectives, we have converged to a common position. We suggest that the authors go through our preamble, which captures this converged view, and also the specific concerns of each reviewer. While some of our major concerns can be addressed by substantive re-writing, cutting down generalisations and over-interpretation others require experimental attention.

Thanks for your inputs. We have worked extensively on the manuscript addressing all the concerns regarding the generalization and over interpretation. However, we are still open to suggestions that might improve the readability of the current manuscript.

One important issue, which was reiterated in our consultations, is exemplified by the following statement and related data "We show that the self-renewing hemocyte progenitors prefer glycolysis while the quiescent progenitors adopt Fatty acid β-oxidation (FAO) for their differentiation. Perturbation of β-oxidation in progenitors results in loss of quiescence…." Though the authors' primary question is centered on distinguishing metabolic states within progenitors, the authors do not identify these. Generic results are interpreted as applying to specific populations and extended to homeostasis.

In the current manuscript, we have focused on the differentiation of progenitors instead of their homeostasis. The stages that we have worked out in detail include the third instar early and late. We want to draw your attention towards the progenitor subsets populating the LG at these two-time points. Please refer to Figure 2M and 2O. The Dome- pre-progenitors are present in early third along with Dome+ and Dome+ Pxn+ Hml+ IP cells. However, in the late third instar, the pre-progenitors (Dome-) are absent. These data are in sync with the findings from (Ferguson and Martinez-Agosto, 2017).

Using Hnf4-Gal4.UAS-GFP as the reporter for β-oxidation, we show that a Dome- preprogenitors start expressing Hnf4.GFP early on, which becomes conspicuous in Dome+ progenitors in the third instar. However, it is downregulated in Dome+Hml+ IP cells (Figure 1 B-C''). A similar trend is noticeable in the quantitative analysis in Figure 1D).

Staging larvae and using appropriate established Gal4 drivers is, of course, acceptable, but the authors have not done this properly. Unlike, for example, the wing and eye discs, the larval LG do not have a uniform, reproducible structure or pattern. One cannot delineate zones morphologically or by using a single marker. Progenitor sub-populations (dome negative pre-progenitors, dome+ progenitors and dome+Pxn+hml+ intermediate progenitors) (reviewed in Banerjee et al., Flybook2019) and the three types of differentiated cells (plasmatocytes, crystal cells and lamellocytes) are functionally and metabolically different. However, the authors do not distinguish between these populations in their analysis.

We have now used appropriate double markers for characterization as well as expression studies and taken utmost care for the staging of the larvae.

For example, mitochondrial analysis is shown at low resolution and in an undefined population (Figure 1L, M)

We understand your concern. We have now performed super-resolution microscopy of the mitochondria in the Dome+ progenitors of late third instar lymph gland (DomeGal4-UASRFP >UAS-mito-GFP). Please refer to Figure 1—figure supplement 1C-C' and Video 1 and discussed in Main text.

and hml>GFP is used only as a marker of differentiation (Figure 4).

Instead of using Hml only, we have now used Dome and P1 to mark the progenitors and differentiated plasmatocytes respectively. Please refer to Figure 4 of the current manuscript.

The assertion of the manuscript is that they are distinguishing between progenitor types, which they call self-renewing and quiescent. This can be done only upon identifying and characterising subpopulations. Further, since they incorrectly use the presence or absence of proliferation marks as a readout for self-renewal and quiescence, it is not clear how the claim that these are two metabolically distinct precursor states is valid. Rather, based on the analysis done (EdU, Figure 1B,C) they are simply proliferating or non-proliferating cells (which may or may not be progenitors), the latter poised for, or in the process of differentiation. Expectedly, their metabolic states differ- and the authors show that these are glycolysis for the former and FAO for the latter. Figure 1B,C: TepIV and EdU labelling – it looks as if there are co-labeled cells, but these are ignored. Also TepIVGAL4 reporter does not label the entire progenitor pool. The existence of self-renewing or quiescent progenitors has not been formally demonstrated in the LG- if the authors had demonstrated their existence, this would have been very interesting indeed.

We are very sorry for the confusion created regarding the cell cycle status of the progenitors. We used the terminologies “self-renewal” and “quiescence” to denote the “proliferative” and “non-proliferative” states of the second instar and late third instar lymph gland progenitors, respectively. We want to bring to your kind attention that a recent work from the lab (Sharma et al., 2019) we found that the Dome+ Pxn- progenitors in the late third instar lymph gland are arrested in the G2 phase of the cell cycle. We have done an extensive analysis of the cell cycle in different progenitor sub-populations to establish the G2 arrest. In the revised version of our manuscript, we have used the terminology “proliferating” and “G2 arrested” to address the cell cycle status of the lymph gland progenitors.

The metabolic fluctuations we are concerned about are intra-LG. This is a major challenge in characterizing metabolic states of heterogeneous progenitors. Since all analysis is done only with one progenitor marker at a time (mainly dome) which is also expressed in early stages of differentiation, or with no marker at all, (Example Figure 1 ; Figure 1—figure supplement 1, compare panels A and C) claims regarding metabolic differences are not substantiated.

We understand your concern and have now used double markers in the current manuscript to assay the metabolic status of the heterogeneous population. Please refer above for the details on progenitor sub populations.

As is also evident from the images in Figure 1, the same tissue from the same genotype and stage varies greatly in terms of size, shape and organisation, hence rigorous quantitation with additional markers is essential.

We fully understand your concern and have taken the utmost care of staging, used appropriate markers, and did rigorous quantitation of expression patterns and phenotypes wherever possible. Thanks for your guidance.

The authors seem confused about what the focus of their study is. They state in the Abstract that the major outcome of the study is coupling FAO with epigenetic modifications linking to regulation of differentiation by JNK, while this part of the study has merit, it is not very relevant to the primary question they pose about blood cell progenitor homeostasis.

Based on your guidance, we have chosen to build up the manuscript concentrating on the role of FAO in progenitor differentiation, which the reviewers, along with you found to have merit. Progenitor homeostasis and glycolysis will be dealt in a separate manuscript at a later date.

The reviews give a very detailed analysis, including comments for almost every panel of Figure 1. However, some examples are expanded on a couple of specific cases below.

Figure 1D, E states that self-renewing early stage progenitors exhibit elevated levels of glucose uptake compared to late-stage quiescent progenitors. In D there is no marker for progenitors or differentiated cells- clearly there are varying intensities of signal. How can one assume that all 2-NBDG cells are progenitors- early or late? Further, given variation between preps, how does one compare levels between second and third instar without having a control to normalise it or assess relative levels between progenitors and differentiated cells? Assuming second and third instar preps were stained and imaged in the same experiment, how much variation is seen? A graph representing this is essential as the lymph gland morphology can vary greatly depending on the status of the larval cultures.

We understand your concern. As mentioned earlier, the current manuscript deals with differentiation, and therefore, we have analyzed third instar stages. Markers for delineating CZ/MZ (dome-Meso-EBFP2) with 2-NBDG have been used along with quantitative analysis. Please refer to Figure 5 B-C'' and Figure 5D for quantitation. The maximized uptake of glucose is in the peripheral hemocytes (2-NBDG in HmlDsRed of control is further endorsed: Figure 5—figure supplement 1)

Additionally, a proliferation assay is required along with the uptake or demarcation of the MZ/CZ boundaries to show reduced uptake in the pro-hemocyte pool of the 3rd instar lymph gland.

We want to draw your attention to the fact that there is a technical constraint in performing any antibody assay or proliferation assay with 2-NBDG. Mild fixation is essential for 2– NBDG assay. Upon prolonged fixation or longer incubation 2-NBDG signals quenches.

In order to map the peripheral signal of 2-NBDG in the LG, we have used Hml-dsRed (Figure 5—figure supplement 1), which makes it evident that the Hml population is indeed high in glucose uptake compared to Dome+ progenitors.

Figure 1F, G – To say that there is high Glut-1 GFP in proliferating progenitors of the second instar, labelling with a proliferation marker, progenitor marker, and differentiation marker is required along with quantitation to show that High Glut 1-GFP is expressed in proliferating progenitors.

Since the current manuscript dwells on FAO mediated differentiation, and hence we plan to include this experiment in a separate manuscript that deals with Glycolysis, progenitor proliferation, and their homeostasis.

Figure 3E-G, the authors claim a higher number of EdU incorporation is seen in the progenitors, however, the images provided show higher EdU signal at the periphery; also CZ/MZ boundary is not demarcated, and progenitor markers are not used.

In the current manuscript, we have brought DomeGFP in the background of whd1 mutant to address this issue. Please refer to Figure 3 G-H and quantitation in Figure 3I.

If the results about FAO and its downstream effects in the LG are de-linked from progenitor sub-types and homeostasis, this would be more acceptable.

Thanks for your guidance. In the current submission, we exactly did this. We have worked on FAO and its downstream effects on the hemocyte progenitor of third instar stages. Our current study connects ROS-JNK and FAO circuit that regulates progenitor differentiation by epigenetic modification.

Currently, two parts of the manuscript (progenitor types and metabolic status connecting to signaling) are not well connected. The authors may want to consider that they are perhaps best dealt separately and not in the same manuscript.

Please refer to the point above. Moreover, we have now strengthened the connection between JNK and FAO with histone acetylation in the current version.

Reviewer #1:

This study by Mandal and colleagues studied the role of metabolic circuitry in stem cell maintenance/differentiation. General impacts of metabolism on cell fates, particularly those between stem cells and their progenies, have increasingly drawn researchers' attention in recent years. By using Drosophila hematopoietic progenitors (in lymph gland), the authors explored the role of metabolic circuitry in this lineage. The main conclusion is that actively proliferating stem cells utilize glycolytic pathway whereas quiescent stem cells favor fatty acid β-oxidation, which promotes differentiation.

Furthermore, they propose that FAO generates acetyl-CoA, which not only generates energy through TCA cycle, but can contribute to protein (e.g. histone) acetylation. Perturbation of FAO leads to defects in quiescence (progenitor starts proliferating and cannot differentiate). Upregulation of FAO leads to stem cell quiescence and precocious differentiation. The authors link this observation to histone acetylation and thus epigenetic regulation.

Figure 1: Does blocking glycolysis have any impact on differentiation? Does it only affect progenitor proliferation without impacting later differentiation?

Although the focus of the current paper is differentiation and β-oxidation, we, however, have done this experiment. Glycolysis inhibition by 2-DG supplemented feeding affects progenitor proliferation without having much effect on Hml+ Cortical Zone. This experiment implicates that progenitor differentiation is not dependent on glycolysis. We intend to put together a separate manuscript on Glycolysis and progenitor homeostasis on a later date.

Figure 2: blocking FAO by means of whd knockdown resulted in increased progenitors and decreased differentiating cells. Further in Figure 3, they show that loss of FAO also increases progenitor proliferation and thus progenitors are not entering quiescence. Collectively these results indicate that FAO is required for progenitors to enter quiescence, and thus their “differentiation defect” is secondary to their inability to enter quiescence. These results are quite clear, but when I was reading the description for Figure 2, I was misled to think that FAO blocks differentiation. Any ways to make this transition (Figure 2 to Figure 3) a bit less misleading?

Thanks lot for appreciating our work. Sorry for the confusion created. We have worked on the description of Figure 2 and modified it.

In Figure 4, they show that upregulation of FAO is sufficient to induce quiescence and subsequent differentiation in progenitors.

Figure 5 demonstrate the inter-relationship between glycolysis and FAO. Defective FAO led to sustained glycolysis.

Figure 6-7 slightly shift in the focus and now addresses FAO's function in histone acetylation. The authors now suggest that it is the key to differentiation.

Figure 8 talks about JNK pathway in progenitor differentiation.

Although the authors' claims are well supported in general, Figures 6-8 make me wonder whether FAO contributes to progenitor cell differentiation through metabolic aspects at all. Although each piece of discoveries in this paper is important, I don't know if the authors can claim that FAO regulates progenitor differentiation through BOTH metabolic regulation and histone acetylation. Especially because Figure7 shows that acetate supplementation is enough to rescue whd mutant phenotype, one has to wonder whether “metabolic” aspect ever contributes to self-renewal/differentiation choices. And if this is all about histone acetylation, what is the importance of glycolysis in self-renewing progenitors? Of course, it is an intriguing possibility that this is nature's elegant solution, natural shifts in metabolism also triggers epigenetic programming. I am not asking to do experiments to tell apart the possibilities, but the fact that their data do not necessarily show the requirement of FAO regulating metabolism per se in regulating differentiation, and the possibilities have to be carefully discussed.

We want to draw the attention of the reviewer to the fact that the metabolic signal ROS is known to trigger JNK in the progenitors to elicit their differentiation (Owusu-Ansah and Banerjee, 2009). We now show that JNK in the progenitors at this stage regulates the metabolic process: FAO through the transcription of whd (the rate-limiting enzyme of FAO).

The end product of FAO is Acetyl CoA, a metabolite implicated in histone acetylation/epigenetic modification. We show that loss of JNK or whd affects progenitor differentiation and histone acetylation. Restoring histone acetylation alone through acetate supplementation results in rescuing the differentiation defect in the mutants.

The acetate supplemented is converted to the end product of FAO: Acetyl CoA, the metabolite that is essential for histone acetylation. The involvement of Acetyl CoA in facilitating differentiation is further evidenced when on genetically downregulating AcCoAs (the major enzyme in acetyl CoA generation) from the progenitor leads to a halt in their differentiation.

Our study thus establishes that for hemocyte progenitor differentiation, the metabolic process FAO involves its metabolite Acetyl CoA for epigenetic modification.

We have discussed this aspect in the Discussion section of the revised manuscript. Thanks for your suggestion.

I understand it might be technically difficult, but can they tell that histones are less acetylated in self-renewing progenitors compared to quiescent progenitors?

Although we are not presenting the observations on the requirement of glycolysis in the proliferating progenitors of early instars, we can comment that there will not be much difference in the overall acetylation of histones. The levels of acetyl COA required for acetylation in proliferating progenitors of the second instar are not compromised owing to the high glycolytic index at this stage. Later on, in development, glycolysis dampens, and FAO picks up to provide the Acetyl COA moiety required for histone acetylation in the progenitors.

I think that Figure 8 is unnecessary. This does not add to the major message of the paper.

Thanks for your input. We have done additional experiments to establish the ROS-JNK-FAO circuit is essential for differentiation. JNK has been previously implicated in epigenetic modification via histone acetylation. Our study unravels the unexpected link of JNK with epigenetics through FAO. We have accordingly modified the model. Please refer to Figure 8—figure supplement 1 of the revised manuscript.

Reviewer #2:

Tiwari et al. present data that they claim suggests a potential role for FAO in metabolic regulation of progenitors in Drosophila larval lymph gland. They interpret their data to say that self-renewing and quiescent progenitors in the LG differ in their metabolic state. The latter exhibit FAO which results in acetylCoA production and thereby increase in histone acetylation and altered gene expression. Pharmacological increase in acetylation has the same effect whereas reduced histone acetylation maintains progenitor self-renewal. They state that the major outcome of the study is coupling FAO with epigenetic modifications linking to regulation of differentiation by JNK.

The metabolic state of blood cell progenitors has been studied extensively in vertebrates and to a limited extent in Drosophila. This study further characterizes progenitor sub-populations for their metabolic status.

However, a major shortcoming of the manuscript is the misinterpretation of literature on which their hypothesis and analyses are based. As a result, the fundamental question being addressed about self-renewal and quiescence in LG progenitors is flawed. Further, there are systemic problems with the methods used to analyze, present and quantify data in this manuscript that make it impossible for me to judge the validity of their conclusions.

We used the terminologies “self-renewal” and “quiescence” to denote the “proliferative” and “non-proliferative” states of the second instar and late third instar lymph gland progenitors, respectively. Our interpretation regarding the quiescent state of the progenitors was based on the published literature available at that time point. We want to bring to your kind attention that in a recent work from the lab (Sharma et al., 2019), we found that the Dome+ Pxn- progenitors in late third instar lymph gland are arrested in the G2 phase of the cell cycle. We have done an extensive analysis of the cell cycle in different progenitor sub-populations to establish the G2 arrest.

However, we do understand that self-renewal has not yet been demonstrated in the LG. We are very sorry for the confusion created regarding the cell cycle status of the early progenitors. In the revised version of our manuscript, we have used the terminology “proliferating” and “G2 arrested” to address the cell cycle status of the lymph gland progenitors.

Controls are rarely presented and cell sub-populations are not identified or marked. It is standard practice to include a control for all gene expression especially when reported for the first time and also to identify and distinguish between cell sub-populations when differences in their properties are claimed. Furthermore, the data in this paper are difficult to interpret and I am confused by how the authors came to their conclusions.

There is a general lack of quantification that would be required to make the interpretations presented.

We fully understand your concern and have taken the utmost care of staging, used appropriate markers, and did rigorous quantitation of expression patterns and phenotypes wherever possible. Thanks for your guidance.

Hence the first part (Figure 1) that emphasizes self-renewal and quiescence distracts from the main findings and the data are unclear and analyses questionable. Since the role of JNK in LG differentiation and its connection to metabolism are already established, the results presented here are only incremental to our understanding of LG progenitor maintenance.

Here we disagree with the reviewer’s statement that our result is only an incremental to our understanding of LG progenitor differentiation. Although ROS dependent JNK activation was implicated in the differentiation of hemocyte progenitors, the mechanistic basis of this phenomenon was not known. We show that JNK can regulate Fat metabolism by the transcriptional activation of the rate-limiting enzyme of FAO, whd. Acetyl-CoA, the product of FAO, in turn, is essential for causing epigenetic modification. Our study, therefore, unravels the unexpected link of JNK with FAO. That FAO activation is mandatory for differentiation is clear from the observation that upon JNK/bsk loss, the halted differentiation can be rescued by upregulating FAO (through L-carnitine supplementation). Please refer to Figure 8 in the revised manuscript.

This is the first report that describes the involvement of the metabolic process FAO and its metabolite Acetyl CoA for epigenetic modification crucial for hemocyte progenitor differentiation. Our study further links the previously described ROS-JNK circuit to metabolism and therefore, should not be undermined.

Further, a lot of the experiments related to this section will have to be repeated with appropriate markers and controls and quantitated rigorously to be convincing. Hence I think the present manuscript is not suitable for eLife.

A detailed review is provided below:

Introduction:

The authors seem to be unaware that of the different hematopoietic populations in vertebrates and how they compare with that of Drosophila. A major problem is the idea that there are self-renewing and quiescent progenitors in the larval LG. It is important to keep in mind that the LG is a powerful but limited model of vertebrate hematopoiesis. Hence a one-on-one correlation between cell types, states and metabolic status should not be made. The Abstract should be corrected to avoid the use of terms such as self-renewal and quiescence. The authors start the Introduction with the aim of testing the relation between metabolic intermediates and histone modifications but deviate to self-renewal and quiescence.

As per the suggestion in the preamble, we have split the paper into two. In the current submission, we have worked on FAO and its downstream effects on the hemocyte progenitor differentiation. We have connected ROS-JNK and FAO circuit regulating progenitor differentiation by epigenetic modification.

Introduction section

Studies on various stem cell scenarios primarily in Hematopoietic Stem Cells(HSCs) have established that various states adopted by stem cells like, quiescence, proliferation, and differentiation are liable to different metabolic requirements.

This suggests that a stem cell can occur in multiple states and is dynamically switching between them. However, this is not the case. Note that the references cited are for multiple sub-populations of vertebrate HSC.

Sorry for the confusion created. We have modified the statement in the revised manuscript.

There has been no formal demonstration of stem cells in the LG nor of self-renewal. However, the authors interpret all their data in the context of self-renewal and quiescence and hence this is incorrect and misleading. Further, the experimental evidence does not support the conclusions (see below).

Although self-renewal has not yet been experimentally demonstrated in LG, it has neither been ruled out. In fact, the stem cells have been identified in larval 1st instar lymph gland through clonal analysis (Minakhina et al., 2010) as well as using several parameters of stem cells like multipotency, niche dependency, lineage tracing, gain and loss of function genetic analyses along with genetic ablation studies (Dey et al., 2016).

Our aim of discussing HSC stems from the idea that the requirement of FAO metabolism is worked out in HSC, but its role in progenitor is yet to described. Using Drosophila larval blood progenitors, we proposed to work out the role of FAO.

Introduction paragraph two

The blood progenitors found in Drosophila late larval lymph gland are akin to the vertebrate common myeloid progenitors (CMP)(Owusu-Ansah and Banerjee, 2009). They are quiescent, have high levels of ROS (Owusu-Ansah and Banerjee, 2009), lack differentiation markers, and can give rise to all Drosophila blood lineages.

This again is a misinterpretation. Firstly, CMPs are not quiescent, only long-term repopulating HSCs are proven to be quiescent.

We have used an understanding prevailing in the field, but understanding your concern we have removed this comparison.

LG progenitor quiescence has not been conclusively demonstrated either. The authors interpret lack of proliferation as quiescence, these are two very different cellular states and associated with different gene expression profiles as well as cell phenotype.

Our interpretation regarding the quiescent state of the progenitors was based on the published literature available at that time point. We want to bring to your kind attention that a recent work from the lab (Sharma et al., 2019), we found that the Dome+ Pxn- progenitors in late third instar lymph gland are arrested in the G2 phase of the cell cycle. We have done an extensive analysis of the cell cycle in different progenitor sub-populations to establish the G2 arrest.

However, we do understand that self-renewal has not yet been demonstrated in the LG. We are very sorry for the confusion created regarding the cell cycle status of the progenitors. In the revised version of our manuscript, we have used the terminology “proliferating” and “G2 arrested” to address the cell cycle status of the lymph gland progenitors.

Finally, unlike in vertebrate studies, the ability of a LG progenitor to give rise to "all Drosophila blood lineages" has not been demonstrated.

We chose to disagree with this point. Studies from Dr. Utpal Banerjee’s laboratory have established that the progenitors are multi-potential (Jung et al., 2005). Also, different studies have shown that transdifferentiation (Leitao et al., 2015) can happen in the blood lineages, causing a change in fate from one cell type to the other (Leitao et al., 2015). We have also tried it out with the potential HSCs as well as the progenitors. In all cases, multipotentcy has been endorsed (Dey et al., 2016).

Introduction paragraph three:

This should be edited to indicate that not all transcription factors are conserved between the two systems.

Modified accordingly.

The authors equate proliferation with self-renewal- these are two different processes that share gene networks. EDU labelling indicates proliferation not self-renewal. Self-renewal is a very special property of stem cells that can also differentiate. Similarly, absence of EDU label indicates the cells are not proliferating or are cycling slowly, quiescence requires existence in the G0 phase, which has not been demonstrated in the LG

We used the terminologies “self-renewal” and “quiescence” to denote the “proliferative” and “non-proliferative” states of the second instar and late third instar lymph gland progenitors, respectively. Our interpretation regarding quiescent state of the progenitors was based on the published literature available at that time point. We want to bring to your kind attention that in a recent work from the lab (Sharma et al., 2019), we found that the Dome+ Pxn- progenitors in late third instar lymph gland are arrested in the G2 phase of the cell cycle. We have done an extensive analysis of the cell cycle in different progenitor sub-populations to establish the G2 arrest.

However, we do understand that self-renewal has not yet been demonstrated in the LG. We are very sorry for the confusion created regarding the cell cycle status of the early progenitors. In the revised version of our manuscript, we have used the terminology “proliferating” and “G2 arrested” to address the cell cycle status of the lymph gland progenitors.

Introduction paragraph four:

Dynamic states of self-renewal, quiescence and differentiation.

Note that a single cell does not switch between these phenotypic states. Differentiation is normally a one-way street down the lineage through each state, and necessitates metabolic shifts.

The data do not provide evidence that the progenitors primed to differentiate rely on FAO- this is merely correlative.

We beg to disagree with this comment. It was previously shown that ROS in the progenitors primes them for differentiation (Owusu-Ansah and Banerjee, 2009). Our results show that despite having high ROS upon FAO loss, the progenitors are unable to differentiate. We have employed molecular as well as genetic analysis encompassing both RNAi mediated as well as the classical loss of function analyses. Additionally, overexpression studies employing genetic constructs as well as pharmacological methods, firmly establishes the central role of FAO in the submitted manuscript.

General comments on results:

Pharmacological treatments: This is another major point that the rest of the paper relies on, yet the authors again show no controls and inadequate quantification.

We had previously incorporated the control for each pharmacological treatment. In the revised manuscript we have quantitated both the controls and experimental. Please refer to figures: Figure 2—figure supplement 1, Figure 4, Figure 5, Figure 7 and Figure 8.

While the entire paper rests on analysis of a few hundred cells in the LG, the Western blot analysis is done with the whole larva and results are interpreted in the context of the LG.

We had done western blot analysis to ascertain the difference, if any, on the level of Histone acetylation in the whole larvae. Once the defect was seen, the same experiment was repeated for acetate-supplemented condition to ensure that, indeed, the supplementation has globally restored the acetylation level.

However, we have not used the above result alone to interpret the result in the context of LG. In addition to several genetic analyses to downregulate FAO, we have done extensive clonal analyses of the lymph gland progenitors. We generated mosaic clones using hsFlp/Ay-GAL4 mediating RNAi knockdown of DrosophilaCPT1 orthologue whd and subjected them to immunohistochemistry using antibody against H3K9 acetylation. The clonal patches positively marked with GFP (where whd is knocked down) showed a significant drop in H3K9 acetylation levels compared to surrounding hemocyte progenitors. Along with H3k9 acetylation, the level of histone H4 acetylation visualized by pan anti-H4 acetylation antibody revealed a drastic drop in total H4 acetylation level in whd knockdown clonal patches. Both the expression of H3K9 and pan H4 acetylation remains unaltered in mock/wild type clones.

Please refer to Figure 6K-M of the revised manuscript.

The authors claim that there are metabolic differences between progenitor sub-populations that correlate with potential to differentiate. However pharmacological treatments such as acetate feeding is done at the level of the whole animal. How much of acetate is sensed by the progenitors in question?

Pharmacological feeding was assessed and standardized in a dose-dependent manner. A lower dosage of 1 mM acetate supplemented fly food was not able to rescue the differentiation defects of whd1. However, a higher dose of 50 mM acetate supplemented fly food was able to rescue the progenitor differentiation as well as histone acetylation levels of the whole larvae.

To assay whether this restores the H3K9 acetylation level in the blood progenitors, we performed anti-immunostaining of H3K9 acetylation and quantitated the mean fluorescence intensity of H3K9 acetylation signal in Dome+ progenitors. Please refer to Figure 7H-L of the revised manuscript.

As evident from this figure that although acetate supplementation has no significant effect on control, it rescues the acetylation level and, therefore, the differentiation defect in whd1 homozygous lymph gland.

Thanks for your suggestion for performing immuno-staining of H3K9 acetylation in situ in the LG.

How does a general increase in acetylation affect histone in various tissues including the LG? As systemic signals are key to maintaining blood progenitor homeostasis, the effects seen could be due to cell extrinsic factors.

Acetate supplementation rescued our differentiation defects both in bsk as well as whd mutant. Our western blot analysis reflected the restoration of histone acetylation in the case of the experiment while there was no change in the control. That this rescue is not systemic becomes further evident when acetate fed control larvae did not show any alteration in differentiation and were comparable to non-fed control. Please refer to Figures 7H-I.

The conclusion that this treatment proves the role of FAO and histone acetylation in progenitor quiescence is far-fetched.

We chose to differ with this statement. Through our extensive molecular and genetic analyses, we have established FAO mediated epigenetic regulation essential for differentiation and not quiescence.

Further there is no evidence of that glycolytic and FAO metabolizing sub-population are differentially affected.

In the current manuscript, we have addressed the differentiation aspect of hemocyte progenitors at third instar larval stages. At the late third instar stage, the LG is populated by Dome+ progenitors only. We have shown that JNK mediated FAO activation is essential for the differentiation of the dome+ progenitors.

It is similarly unclear in which progenitor sub-populations whd and JNK signaling are affected.

Sorry for the confusion caused. JNK and whd signaling are affected in the Dome+ progenitors of late third instar LG.

Quantification is required for differentiation checked by P1 and progenitor status should be checked by dome; also differentiation to crystal cells and lamellocytes should be examined.

Thanks for your suggestions. We have included all these analyses in the revised manuscript. Please refer to Figure 2. However, as far as lamellocyte is concerned, neither wild type nor whd1 mutant lymph glands have them. We have not included this result.

This is the first report of metabolic requirement for progenitor maintenance; a thorough analysis of various differentiation markers is essential.

Thanks for recognizing our effort. In the revised manuscript we have tried our best to strengthen our findings.

Figure wise comments

Figure 1 is fraught with misinterpretations. (Results paragraph one and two). Further there is no quantitation whatsoever and a lack of controls. Two different developmental stages are compared and cell identity is arbitrarily assigned. No markers are used for self-renewing or quiescent cells and it is not clear how the authors identify these. Edu positive cells are only a subset of the second instar LG (these could be differentiating) and seen only in the CZ of the third instar.

In the current version we have used relevant markers to characterize the phenotype.

Figure 1D, E states that self-renewing early stage progenitors exhibit elevated levels of glucose uptake compared to late stage quiescent progenitors. In this case a proliferation assay is required along with markers to demarcate the MZ/CZ boundaries to show that the uptake is less in the prohemocyte pool of the 3rd instar lymph gland.

We want to draw your attention to the fact that there is a technical constraint in performing any antibody assay or proliferation assay with 2-NBDG. Mild fixation is essential for 2 – NBDG assay. Upon normal fixation or more extended washing (required for Antibody staining) 2-NBDG signal quenches. The restricted peripheral signal of 2-NBDG in HmldsRed suggests that the maximum uptake is in the peripheral hemocytes instead of the inner ones (Figure 5—figure supplement 1).

It is likely that the 2NBDG labeled cells could be late progenitors in the second instar that persist in CZ. Without appropriate markers this cannot be distinguished. Further third instar progenitors may be slower proliferating or arrested in the cell cycle, this is not quiescence. A cell cycle analysis needs to be done through L2 and L3 to resolve the two states, so called self-renewal and differentiation.

What the authors call quiescence is likely a non-proliferative state that precedes differentiation, which is well documented in many developmental contexts. Similarly as mentioned in Results paragraph two, glycolysis is seen in rapidly proliferating cells, such as cancer cells, this is not necessarily self renewal.

GlutGFP expression is patchy (Figure 1F, G) and not clear whether it is in or around the cells, this looks more like background. A no primary antibody control and a Glut RNAi control are required. MZ should be marked by dome or DEcad and perform EdU labeling in order to show that High Glut 1-GFP is expressed in proliferating progenitors. Similarly aldolase expression (supplement 1A, B) is not convincing and is seen only in the periphery of the second instar.

We thank the reviewer for the input. However, in the current manuscript, the focus is on progenitor differentiation and the requirement of FAO in this process. We intend to communicate the significance of glycolysis and early hemocyte progenitor homeostasis in a different manuscript and include the results in it.

Figure 1K: Which area of the LG is shown? Nuclear staining should be included.

Nuclear staining has now been included for most of the figures in the revised version.

Figure 1L-O' it is impossible to comment on the status of mitochondrial network with this analysis. Which part of the LG is shown? High resolution images with live tracking and video microscopy are required to analyze mitochondrial morphology, length and dynamics. Without detailed information regarding these parameters in L2 and L3 progenitors and in differentiated CZ cells, one cannot make any comparisons as there is a large variation in mitochondrial size and morphology within any population. This needs co-analysis with progenitor and differentiation markers and thorough quantitation- CZ/ MZ mitochondrial pattern needs to be mapped first.

The focus of our current submission is FAO’s role in progenitor differentiation. We have concentrated on L3. Since the site of FAO is mitochondria, we wanted to have a feel for the mitochondrial status in the Dome+ progenitors. In the current manuscript, we have performed super-resolution microscopy to visualize the mitochondria better, the genotype used is DomeGal4.UAS-RFP >UAS-mito-GFP. Please refer to Figure 1—figure supplement 1 and Video 1.

Figure 1Q, R: the accumulation/increase of Lsd-2 in the 3rd instar lymph gland would be clear if 2nd instar images are presented for Nile Red.

Nile red data have now been endorsed by LipiTOX (a validated assay for neutral lipids). All these results have been quantified and presented in Figure 1—figure supplement 2A-C.

Figure 1 requires quantification and insets indicating region of interest; double labeling with progenitor marker and or CZ markers or EdU labeling is essential.

We have addressed all these issues in the current Figure 1.

Figure 2: Dapi staining or demarcation of LG boundary as well as MZ/CZ marker staining should be shown. Crystal cell and lamellocytes differentiation status should also be analyzed.

We have included the DAPI channel and the quantitation data for this experiment. Please refer to Figure 2.

2H,L Q Graph is misleading, it suggests there is a decrease in dome+ area. Y axis should indicate plamatocyte % or area. What is differentiation index?

We have modified the Y-axis to cortical zone area by total lymph gland area in Figure 2.

Figure 3E-G, the authors claim higher number of EdU incorporation is seen in the progenitors, however the images provided show higher EdU signal at the periphery; also CZ/MZ boundary is not demarcated, progenitor marker should be used and data quantitated.

We have performed the same experiment with markers for the progenitor zone (Dome+). Quantitative analysis has now been included for the same to resolve the confusion.

Figure 5F-H quantitation required for EdU labeling to show that feeding with 2-DG rescues the hyperproliferation phenotype.

We have provided the required quantitation for this experiment. Please refer to Figure 5I-L.

Figure 6A-C claims loss of Histone Acetyl Transferase (HAT), Gcn5 and chm phenocopies differentiation defect seen in the progenitors of FAO loss of function. What is the effect on proliferation in panels 6A-G. Does it corroborate with active proliferation seen in whd mutants.

We thank the reviewer for the suggestion. We have performed the EdU incorporation in Dome-GAL4-UAS-GFP> UAS-Gcn5-RNAi and Dome-GAL4-UAS-GFP> UAS-chm-RNAi. In both, the case unlike whd1 mutant, the hemocytes were not proliferating. We can infer that Gcn5 and chm facilitate hemocyte progenitor differentiation.

For Figure 7F materials method suggests western blots were done from larva of the respective genotypes, this cannot be correlated to the LG effects. It would be more appropriate to show H3K acetylation in situ in the lymph gland.

Thanks for your comments. We have now provided evidences of the upregulation of H3K9 acetylation level in the lymph gland by immunostaining in both Acetate as well as L-carnitine fed conditions.

Please refer to Figure 7H-L and the main text of the revised manuscript for acetate supplementation. For L-carnitine supplementation, please refer to Figure 7O-Q for H3K acetylation status in situ in the lymph gland.

Figure 8A- C, expression of dominant negative allele of basket (bsk, Drosophila ortholog of Jun-Kinase) results in a compromised differentiation analogous to whd knockdown. Can the phenotype be rescued by feeding L-carnitine or acetate?

We thank the reviewer for this suggestion.

We have done both acetate as well as L-carnitine supplementation Dome> bskDN.

Indeed in both cases, we see a significant rescue of differentiation. Please refer to (Figure 8M-Q) for acetate feeding and L-carnitine supplementation (Figure 8R-V).

These results are discussed in in the main text of the revised manuscript

The effect of Jun (bsk) on whd is shown at the transcriptional level. Does Jun (bsk) affect enzymes of the FAO pathway?

Thank you very much for this suggestion.

We have performed qRT-PCR analysis of transcripts of enzymes involved in FAO pathway in Dome-GAL4- UAS-GFP> UAS-BskDN. The transcription of CPT1/whd, Mcad, Mtpa, Scully, Mtpb, and yip2 were assayed upon down-regulation of bsk from hemocyte progenitors. The transcript level of CPT1/whd indicated a ~41% drop while the expression of other b-oxidation enzymes showed no significant alteration. This observation established that JNK controls the transcription of CPT1/whd, thereby regulating FAO. Please refer to Figure 8I and the main text of the revised manuscript.

Also Jnk is known to activate Foxo, what is the status of Foxo reporter (Thor-LacZ) in whd mutants.

Due to the technical limitation of recombining thor lacZ with whd1 (both are on the second chromosome), we have used another bona fide reporter for JNK: puckered (puc)lacZ. Our analysis reveals that in the whd1 mutant lymph gland, there is a significant increase in puc lacZ expressing cells indicating a higher JNK activation in the progenitors. It therefore is evident that in whd1 LG high ROS and high JNK are unable to initiate differentiation in absence of FAO.

Author response image 1

Foxo overexpression advances differentiation (Owusu-Ansah and Banerjee, 2009), does overexpression of Foxo affect/ rescue whd mutant phenotype by inducing differentiation.

Thanks for your suggestion. We have done these experiments. Upon overexpression of Foxo, we also see a similar increase in differentiation, as reported (Owusu-Ansah and Banerjee, 2009). However, one copy loss of whd in the background of Foxo overexpression rescues the ectopic differentiation. Please refer to Figure 8D-H and the main text of the revised manuscript.

[Editors’ note: what follows is the authors’ response to the second round of review.]

Essential revisions:

Authors should talk to non-Drosophila colleagues to help them in thoroughly revising the following aspects of the manuscript:

1) Abstract is not well written and should be improved, specially the last two sentences. It does not summarize the major aspects of the work and will not attract the attention of any reader that has not a special interest in the Drosophila LG.

Thanks for your suggestion. We have accordingly worked on the Abstract as well as the Introduction and made necessary changes to attract the attention of the readers.

2) Cartoon depicted in Figure 1A and explanation in the Introduction do not have anything to do with each other (CZ only explained in the text and others only in the figure) in the figure and does not clarify anything. It's just confusing and should be completely changed and re-done.

We have worked on Figure 1A and also on the explanation of the same. Please refer to Introduction paragraph three.

3) Last paragraph of the Introduction should summarize the major findings and they do not so. References to the JNK/ROS axis are lacking. Jnk itself is not introduced. Again, this part of the manuscript is intended to attract the non-expert community

Please refer to the revised version of the manuscript final paragraph of the Introduction. We have built up the last paragraph keeping the suggestions in mind.

Thanks for your input.

4) First part of the results: the why of using two developmental stages and different markers (pvf2, Dome, Hml, etc) is not clearly explained anywhere.

Drosophila larval lymph gland consists of heterogeneous progenitors. Using molecular markers, these progenitor subpopulations are referred to as pre-progenitors, progenitors, and intermediate progenitors, respectively (Figure 1). The pre-progenitors can also be visualized by Pvf2 expression in first, second, and early third instar larval lymph gland. However, by the late third instar, only progenitors and intermediate progenitors are present in the lymph gland. Therefore, it was necessary to use an early third instar stage to characterize the pre-progenitors. We have explained this in the revised manuscript in the first paragraph of the Results section.

Dome+, Dome+ Pxn+, Dome-, etc nomenclature is confusing (2N-O). Any way of improving it? Authors should be able to avoid technicalities.

We have used the published annotation to avoid confusion. However, instead of using these nomenclatures throughout, in the revised manuscript, we have once described it and thereafter described them without the technical details. “These progenitor subpopulations will be henceforth referred to as pre-progenitors, progenitors, and intermediate progenitors (IPs), respectively.”

5) Tens of mistakes (prolifeartes, progenitors experiences, etc) should be corrected.

We are extremely sorry for these unintentional mistakes. We have tried to fix all of them. Thanks for pointing them out.

6) Writing and conclusions in all sections should be improved.

Thanks for your input. We have worked extensively on the writing and conclusions in all the sections keeping the reviewers suggestions in mind.

Scientific issues:

7) Lipid markers and CG3902 expression. A Z section is required to demonstrate that the differential expression/incorporation levels between progenitor and differentiated cells is real not a consequence of the focal plane or thickness of the tissue.

We want to draw your attention for the mentioned figures. Five optical sections of 1µm thickness from the middle of the Z stack were merged into a single section. We have mentioned this detail in the respective figure legends.

Therefore, the differential expression/incorporation levels observed between progenitor and differentiated cells evident are not an outcome of the focal plane or thickness of the tissue.

Please find the montage of the confocal stacks from which the Z projection was used for the panel for your information.

Author response image 2
CG3902 YFP.

Five optical sections of 1µm thickness from the middle of the Z stack (No. 9-13) were merged into a single section for this panel.

Author response image 3
LipidTOX.

Five optical sections of 1µm thickness from the middle of the Z stack (No. 7-11) were merged into a single section for this panel.

Author response image 4
Nile Red.

Five optical sections of 1µm thickness from the middle of the Z stack (No. 10-14) were merged into a single section for this panel.

8) Authors should use ubiquitous Gal4 drivers (NOT dome-Gal4) to demonstrate differences in the amount of mitochondria and cell cycle profiles (Fucci) between progenitor and differentiated cells.

In the first version of the manuscript, we did use the Ubi-Gal4 to drive UAS-mitoGFP. However, we were suggested in the first review to use Dome gal4 instead. Since we were characterizing the dome progenitors, we went ahead and did dome-Gal4>UAS-mito-HA-GFP or Dual FUCCI.

Also, to be noted for reasons unknown, Ubi-Gal4 is not uniformly expressed in the lymph gland.

Author response image 5

Since we have already analyzed the progenitor pool, the best we could do was to look at the differentiated cells to have a complete picture. Taking the suggestion in mind, we used Pxn-Gal4 to drive UAS-mito-HA-GFP and UAS-FUCCI to assay the mitochondria and cell cycle status of the differentiating cells. Since the differentiated cells are not the focus of the current study, we have not included these data in the revised manuscript. However, we would like to share our findings with the reviewers.

Author response image 6

This experiment reveals that there are plenty of reticular mitochondria in the differentiating hemocytes. Despite having plenty of mitochondria, our results demonstrate a higher glucose uptake in the differentiating hemocytes with basal level expressions of the components of FAO. These results indicate that glucose metabolism followed by oxidative phosphorylation might be prevalent in the differentiating cells. It will be interesting to pursue metabolic regulation of differentiating cells as a separate project.

Author response image 7

This experiment demonstrates that in the late third instar lymph gland most of the differentiating cells are in S phase (red), very few in G1 (green) and G2-M(yellow).

9) Some negative results (changes in Hh and ROS levels w/o any impact on the process) should be moved to supplementary data.

We have moved these two data in the Supplementary Figure. Please refer to: Figure 3—figure supplement 1. High ROS is associated with the differentiation of the dome+ progenitor (Owusu-Ansah and Banerjee, 2009). However, we found that the progenitors of whd1 despite having high levels of ROS are unable to differentiate. At this point, we started appreciating the role of FAO in differentiation. Keeping this twist in mind, although we have moved this data in a supplementary figure, we chose to club it under separate heading to emphasize the importance of FAO in the differentiation process.

Please refer to subsection “Failure in differentiation of hemocyte progenitors upon loss of FAO is not due to decline in ROS levels”.

10) CycB degradation of FlyFucci is not degraded in G1, as authors claim in subsection “Loss of FAO caused an increase in progenitor proliferation, and high redox levels”.

Our sincere apology for this mistake, thanks a lot for bringing this to our notice. We have corrected it in the revised version.

11) Single channels to monitor differences in EdU incorporation levels should be included (3G-I, 5M-P). Figure 5E, F is not described properly in the text.

We have included the single channel for EdU incorporations in all the mentioned figures. Please refer to Figure 3A-B', Figure 4 I-J', Figure4 L-M', and Figure 5 I-L' of the revised manuscript. We have described Figure 5E, F in subsection “FAO loss in hemocyte progenitors leads to sustained glycolysis”.

12) Y axis in panel 3F: should be ROS and not Ci, right?

Sorry for this unintentional mistake. Thanks for pointing this out. We have corrected it in the revised version.

13) Materials and methods: Codes and/or references of all Fly stocks required. By the way, many of the fly lines used in the work are not included in Materials and methods. Stocks such as AyGal4 should be properly explained.

We have taken care of all of these in the revised version. Please also refer to the Materials and methods for details of stocks generated for this study. AyGal4 is now described in the Materials and method section in details.

14) There is a lot of room for improvement in the organization of panels within Figures. Labels of most histograms are too small and error bars are not explained in the legends (SD or SDM?).

Thanks for your suggestions. We have worked on them and are open to any further suggestions. We have increased font size of the labels of all the histograms and have now explained the error bar as SDM in the figure legends and Materials and methods section.

15) What is the trigger for differentiation? Is dietary lipid content a regulator of this as has been shown in vertebrates for regulation by FAO?

Mere carnitine supplementation is not sufficient as carnitine is not an essential nutrient, is abundant in the body under regular diet and carnitine deficiency results from mutations in its transport machinery. As the claim is that whd links JNK to FAO, and is transcriptionally regulated by JNK, a genetic rescue experiment with whd is essential, especially as Foxo overexpression increases differentiation. This manuscript often relies on compound treatment strategies alone despite having a genetically tractable model. Yet the role and effect of endogenously generated systemic cues is completely disregarded.

We want to draw your attention to a genetic manipulation that we had done in the previously submitted manuscript (Figure 8D-H). Genetic removal of one copy of whd was sufficient to prevent the precocious differentiation observed in progenitor specific overexpression of FOXO. To endorse our claim further in the current manuscript, we done a converse experiment. We have overexpressed whd in the progenitors, which lacked JNK (dome-Gal4. UAS-GFP>UAS-bskDN, whd-OE). Our results show that the progenitors that were unable to differentiate upon loss of JNK, resume their differentiation, with whd overexpression. Please see Figure 8—figure supplement 1L-P.

From both the genetic data we can conclude that FAO acts downstream of JNK to facilitate differentiation. Please refer to paragraph six of subsection “JNK signaling regulates FAO in the hemocyte progenitors”.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Essential revisions

1) In the Introduction, the authors contrast hematopoietic stem cells (HSCs) in mammals with Drosophila hematopoietic progenitors, assuming that studying Drosophila hematopoietic progenitors will provide insights into how mammalian non-HSC progenitors are regulated by FAO. However, given that Drosophila does not have HSCs, and whether they represent HSCs or progenitors remain unclear (or whether it's comparison to mammalians HSC vs. progenitor is more relevant). In mammalian HSCs, FAO is required for stem cell maintenance, whereas in Drosophila LG, FAO is required for differentiation. I see that this difference may have driven the authors to make the above contrast, but I don't think it's a well-grounded argument and they should modify the Discussion accordingly.

We have modified the Introduction and Discussion accordingly in the revised version.

2) Figure 1—figure supplement 1C shows reticular mitochondria as an indication of high FAO. Ideally, they should show less reticular mitochondria in other cell types in comparison.

We are happy to let you know that we had this data for the differential mitochondrial structure in differentiated hemocytes compared to progenitors (Streptavidin marking the mitochondria in Hml >GFP lymph gland). For reasons unknown to us HmlΔ-GAL4 is unable to drive UAS-mito-HA-GFP, which we had employed to assay the status of mitochondria in the progenitors. We, had therefore, used streptavidin labeling to visualize the mitochondrial status in the differentiated hemocytes. Interestingly, differentiated hemocytes (HmlΔ-GAL4> UAS-GFP) show less reticular mitochondria, in comparison to the progenitors (Hml>GFP negative). We have included this result in the revised version (please refer to Figure 1—figure supplement 1D-D').

3) Figure 4: to nail down the pathway, the epistasis analysis would be ideal, such as combination of genetic mutations (inhibit differentiation, such as whd mutant) and L-carnitine (drive differentiation). Especially, they conduct similar experiments in Figure 5, Figure 6, and Figure 8.

Although this could have been an ideal experiment, we did not carry out this analysis previously. Since whd null mutant and carnitine, both are part of the carnitine shuttle that imports carnitine fat moiety into mitochondria. CPT1/whd is the enzyme present in the outer mitochondrial membrane, which converts fatty acyl-CoA into fatty acyl-carnitine by utilizing the carnitine moiety. Therefore, we rationalized that upregulation of FAO by carnitine supplementation in whd null homozygous mutants would not aid in the epistatic analysis.

However, we have now used whd1 heterozygous, which incidentally also gives an impressive phenotype (more progenitor, less differentiation compared to control). L-carnitine feeding to the heterozygous restores differentiation to a great extent. We have included this in the revised version (please refer to Figure4—figure supplement 1A-E).

4) Figure 5I-L, 2-DG treatment rescues G2/M arrest defects in whd mutant, but does not seem to rescue differentiation defect. Any discussions?

We had mentioned this aspect in the Result section. Now in the revised version we have expanded this thought and included it in the Discussion, as “The glycolytic surge in FAO mutants is not capable to take them through the differentiation process. This indicates that for the process of differentiation the acetyl moiety derived from FAO plays a key role to facilitate hemocyte progenitor differentiation”.

https://doi.org/10.7554/eLife.53247.sa2

Article and author information

Author details

  1. Satish Kumar Tiwari

    Developmental Genetics Laboratory, Department of Biological Sciences, Indian Institute of Science Education and Research (IISER) Mohali, Mohali, India
    Contribution
    Conceptualization, Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing - original draft, Writing - review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-1827-4603
  2. Ashish Ganeshlalji Toshniwal

    Molecular Cell and Developmental Biology Laboratory, Department of Biological Sciences, Indian Institute of Science Education and Research (IISER) Mohali, Mohali, India
    Present address
    University of Utah, Department of Biochemistry, Salt Lake City, United States
    Contribution
    Formal analysis, Validation, Investigation, Visualization
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-8957-7970
  3. Sudip Mandal

    Molecular Cell and Developmental Biology Laboratory, Department of Biological Sciences, Indian Institute of Science Education and Research (IISER) Mohali, Mohali, India
    Contribution
    Formal analysis, Methodology, Writing - review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-2211-483X
  4. Lolitika Mandal

    Developmental Genetics Laboratory, Department of Biological Sciences, Indian Institute of Science Education and Research (IISER) Mohali, Mohali, India
    Contribution
    Conceptualization, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing - original draft, Project administration, Writing - review and editing
    For correspondence
    lolitika@iisermohali.ac.in
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-7711-6090

Funding

Wellcome Trust/DBT India Alliance (IA/S/17/1/503100)

  • Lolitika Mandal

CSIR

  • Satish Kumar Tiwari

Indian Institute of Science Education and Research Mohali

  • Lolitika Mandal
  • Sudip Mandal

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

Acknowledgements

We thank I Ando, U Banerjee and M Kanost for reagents. We thank all members of the two labs for their valuable inputs. Thanks to Parvathy Ramesh for her help in imaging with Olympus Flouview FV10i. We thank IISER Mohali’s Confocal Facility, Bloomington Drosophila Stock Center, at Indiana University, DGRC (Kyoto), VDRC (Vienna) and Developmental Studies Hybridoma Bank, University of Iowa for flies and antibodies. Models ‘Created with BioRender.com’. DBT/Wellcome-Trust India Alliance Senior Fellowship [IA/S/17/1/503100] to LM and Institutional support to SM and CSIR funding to SKT for this study duly acknowledged.

Senior and Reviewing Editor

  1. K VijayRaghavan, National Centre for Biological Sciences, Tata Institute of Fundamental Research, India

Reviewer

  1. Yukiko M Yamashita, University of Michigan, United States

Publication history

  1. Received: November 1, 2019
  2. Accepted: June 11, 2020
  3. Accepted Manuscript published: June 12, 2020 (version 1)
  4. Version of Record published: July 9, 2020 (version 2)

Copyright

© 2020, Tiwari 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

  • 2,755
    Page views
  • 331
    Downloads
  • 0
    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)

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

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

Further reading

    1. Cell Biology
    2. Developmental Biology
    Delia Ricolo, Sofia J Araujo
    Research Article

    Subcellular lumen formation by single-cells involves complex cytoskeletal remodelling. We have previously shown that centrosomes are key players in the initiation of subcellular lumen formation in Drosophila melanogaster, but not much is known on the what leads to the growth of these subcellular luminal branches or makes them progress through a particular trajectory within the cytoplasm. Here, we have identified that the spectraplakin Short-stop (Shot) promotes the crosstalk between MTs and actin, which leads to the extension and guidance of the subcellular lumen within the Terminal Cell (TC) cytoplasm. Shot is enriched in cells undergoing the initial steps of subcellular branching as a direct response to FGF signalling. An excess of Shot induces ectopic acentrosomal luminal branching points in the embryonic and larval tracheal TC leading to cells with extra subcellular lumina. These data provide the first evidence for a role for spectraplakins in single-cell lumen formation and branching.

    1. Developmental Biology
    Shashank Gandhi et al.
    Research Article Updated

    During gastrulation, neural crest cells are specified at the neural plate border, as characterized by Pax7 expression. Using single-cell RNA sequencing coupled with high-resolution in situ hybridization to identify novel transcriptional regulators, we show that chromatin remodeler Hmga1 is highly expressed prior to specification and maintained in migrating chick neural crest cells. Temporally controlled CRISPR-Cas9-mediated knockouts uncovered two distinct functions of Hmga1 in neural crest development. At the neural plate border, Hmga1 regulates Pax7-dependent neural crest lineage specification. At premigratory stages, a second role manifests where Hmga1 loss reduces cranial crest emigration from the dorsal neural tube independent of Pax7. Interestingly, this is rescued by stabilized ß-catenin, thus implicating Hmga1 as a canonical Wnt activator. Together, our results show that Hmga1 functions in a bimodal manner during neural crest development to regulate specification at the neural plate border, and subsequent emigration from the neural tube via canonical Wnt signaling.