Genetic interactions regulate hypoxia tolerance conferred by activating Notch in excitatory amino acid transporter 1-positive glial cells in Drosophila melanogaster

Abstract Hypoxia is a critical pathological element in many human diseases, including ischemic stroke, myocardial infarction, and solid tumors. Of particular significance and interest of ours are the cellular and molecular mechanisms that underlie susceptibility or tolerance to low O2. Previous studies have demonstrated that Notch signaling pathway regulates hypoxia tolerance in both Drosophila melanogaster and humans. However, the mechanisms mediating Notch-conferred hypoxia tolerance are largely unknown. In this study, we delineate the evolutionarily conserved mechanisms underlying this hypoxia tolerant phenotype. We determined the role of a group of conserved genes that were obtained from a comparative genomic analysis of hypoxia-tolerant D.melanogaster populations and human highlanders living at the high-altitude regions of the world (Tibetans, Ethiopians, and Andeans). We developed a novel dual-UAS/Gal4 system that allows us to activate Notch signaling in the Eaat1-positive glial cells, which remarkably enhances hypoxia tolerance in D.melanogaster, and, simultaneously, knock down a candidate gene in the same set of glial cells. Using this system, we discovered that the interactions between Notch signaling and bnl (fibroblast growth factor), croc (forkhead transcription factor C), or Mkk4 (mitogen-activated protein kinase kinase 4) are important for hypoxia tolerance, at least in part, through regulating neuronal development and survival under hypoxic conditions. Becausethese genetic mechanisms are evolutionarily conserved, this group of genes may serve as novel targets for developing therapeutic strategies and have a strong potential to be translated to humans to treat/prevent hypoxia-related diseases.


Introduction
Hypoxia (O 2 deprivation) is a common pathological factor in many human diseases, including apnea of prematurity, hypoxemia in ICU settings resulting from neurologic, hematologic, respiratory and cardiac diseases, stroke, and cancer (Sands and Owens 2015;Adler et al. 2017;Schmidt et al. 2017Schmidt et al. , 2019Group 2018;Thille et al. 2018;Di Fiore et al. 2019). To date, strategies to treat or prevent hypoxia-induced injury are very limited. Thus, understanding the mechanisms regulating tolerance or susceptibility to hypoxia is crucial for developing effective therapeutic strategies.
A number of studies have demonstrated that Notch signaling plays an important role in regulating the cellular and molecular mechanisms underlying the responses to hypoxic stress (for selected reviews, see Andersson et al. 2011;Alberi et al. 2013;Marignol et al. 2013;Zhou and Haddad 2013;Borggrefe et al. 2016;Fearon et al. 2016;Arumugam et al. 2018). For example, (1) it has been shown that Notch intracellular domain coordinates with HIF signaling to regulate cellular response to hypoxia (Gustafsson et al. 2005;Pear and Simon 2005;Sainson and Harris 2006); (2) we have previously discovered that Notch signaling is activated and plays a critical role in regulating hypoxia tolerance in Drosophila melanogaster (i.e., D.melanogaster carrying Notch loss-of-function alleles are super-sensitive to low O 2 , and, in contrast, the fruit flies carrying gain-of-function alleles are remarkably resistant) (Zhou et al. 2008 (3) neuronal-or glial-specific activation of Notch rescues naïve flies from lethal degree of O 2 deprivation ; and (4) Notch signaling is an evolutionarily conserved mechanism regulating adaptation to low O 2 environments not only in Drosophila but also in humans (Jha et al. 2016). However, the molecular mechanisms underlying the role of Notch signaling in regulating hypoxia responses are still not well understood. BecauseNotch is highly pleiotropic, it is clear that specific downstream responses to Notch activation depend on cellular context and its synergic integration with other signaling pathways (Bray and Bernard 2010;Ho et al. 2018). Furthermore, due to the broad physiological role of Notch signaling in cell differentiation and metabolism, therapeutic strategies directly targeting Notch may introduce severe side-effects (Rizzo et al. 2014;Takebe et al. 2014;Mollen et al. 2018). Therefore, identifying genetic modifiers or downstream effectors that mediate the role of Notch in hypoxia are very important for developing effective therapeutic strategies.
It has long been recognized that hypoxia does not affect all organs/tissues of the body equally. Of all the organs, brain is one of the organs that are very sensitive to oxygen deprivation (Leach and Treacher 1998;Purins et al. 2012). We previously discovered that hypoxia-induced lethality can be rescued by activating Notch signaling only in the excitatory amino acid transporter 1 (Eaat1)-positive glial cells in D.melanogaster, which is likely through cell injury prevention in the central nervous system. It is remarkable that the activation of Notch in this group of glial cells alone is sufficient to enhance the organismal survival under prolonged hypoxic conditions . Presumably, this group of Eaat1-positive glial cells regulates hypoxia tolerance through suppression of glutamate neurotoxicity, neuronal proliferation as well as the growth and organization of axons (Gibson 1989;Ikonomidou et al. 1989;Stacey et al. 2010). Furthermore, since we have identified a group of evolutionarily conserved genes and biological processes (including Notch signaling) between D.melanogaster and human highlanders (Jha et al. 2016) that are important for survival in hypoxic environment, some of these conserved genes may interact with Notch signaling to regulate Notch activation-conferred hypoxia tolerance. In this study, we studied the genetic interactions between Notch signaling and a group of these conserved genes [i.e.,branchless (bnl), crocodile (croc), Epidermal growth factor receptor (Egfr), grain (grn), hairy (h), invected (inv), and MAP kinase kinase 4 (Mkk4)]and identified modifiers that regulate Notch activation-conferred hypoxia tolerance in D.melanogaster.

Materials and methods
Drosophila stocks and culture conditions  (Go et al. 1998), the w 1118 stock (FBst0003605) was used as one of the negative controls in the dual-UAS/Gal4 experiments.
Flies were maintained in vials with Cornmeal-Molasses-Yeast medium. Flies for the hypoxia tolerance assay were prepared as follows. 20 Females of UAS-RNAi were crossed with 20 males of EN line, Eaat1-Gal4 or da-Gal4 and incubated at room air/temperature condition. After 24 hours, parents were removed. One group of the vials (n ¼ 3 to 6 vials) containing the embryos were transferred to a computerized atmosphere chamber supplied with 5% oxygen for hypoxia treatment, and the other group of vials (n ¼ 3 to 6 vials) were cultured in room air condition and used as controls.

Hypoxia treatment and survival test
Three-to five-day-old Eaat1-Gal4and EN males (n ¼ 10) were crossed to the UAS-RNAi virgin females (n ¼ 10) targeting specific genes. Sufficient time was given ($3 days) for the flies to mate/ cross. Simultaneously the w 1118 and the Gal4 line were 'selfcrossed' and used as negative controls, the [EN] x [w 1118 ] cross was used as a positive control. Each set of crosses were in 3 to 6 replicate vials. The vials were kept under ambient conditions for 48 hours so that the flies can lay sufficient number (50 to 100) of eggs. After 48 hours, the adults were transferred to a new vial. For the hypoxia tolerance test the original vials were then transferred to a computer-controlled hypoxia chamber, constantly maintained at 5% oxygen. The chambers were in a room with 12/ 12 hours light/dark cycle, and temperature $22 C. The adults from the new vials (i.e., from the second batch of vials) were discarded after 48 hours, and the vials with the eggs were kept at ambient oxygen conditions ($21% O 2 ) also with 12/12 hours light/ dark cycle (temperature $22 C) as room air controls. After 21 days, the ratio of the empty pupae (eclosed) to the total number of pupae formed (eclosed þ uneclosed) in each vial was calculated to determine the eclosion rate. Due to the fact that most of the collected embryos will be at an early larval stage, this hypoxia test was mainly to determine the survival rate from larval stage to adulthood, as representated by eclosion rate.

Immunostaining and microscopy
The third instar larval brain samples for immunostaining were prepared according to previous descriptions (Wu and Luo 2006;Zhou et al. 2011). Briefly, brains of wandering third instar larvae were dissected in PBS and fixed in 4% paraformaldehyde in PBS. Cell membranes were permeabilized with 0.3% Triton X-100 in PBS, blocked with 7% goat serum, and put in primary antibody overnight at 4 C (or for 1 hour at room temperature). This procedure was followed by washes in 0.3% Triton X-100 in PBS, incubation with secondary antibody for 90 minutes. Following staining, the brain samples were washed multiple times with PBS, mounting and microscopy. The mouse anti-NICD (undiluted in-house supernatant) and rat anti-elav (1:50) were obtained from the Developmental Studies Hybridoma Bank (DSHB); rabbit anti-repo (1:500) was a kind gift of Dr. G. Technau at the Institute of Genetics, University of Mainz, Mainz, Germany. Secondary antibodies used were goat anti-mouse; goat anti-rabbit; and goat anti-rat conjugated to Alexa 488, 546, or 647 (1:250, Invitrogen). The Prolong Gold anti-fade reagent with DAPI (Invitrogen) was used as mounting media. Confocal microscopy was performed in the University of California at San Diego Neuroscience Microscopy Shared Facility. Imaging was done on a confocal microscope (Olympus FV1000) and the images were processed with Image J.

Statistical analysis
Data were analyzed and graphed using GraphPad Prism 6 software.

Results and discussion
The Eaat1-positive cells are a group of glia regulating glutamate metabolism and transport as well as neuronal activity in Drosophila brain (Rival et al. 2004;Peco et al. 2016). We have previously discovered that hypoxia activates Notch signaling (Zhou et al. 2008), and this activation particularly in the Eaat1-positive glial cells can significantly enhance survival of Drosophila in severely hypoxic environments . We aimed in this study to dissect the potential mediators of Notch-conferred hypoxia tolerance. , and third (C) instar larval as well as adult (D) brain. Colocalization of GFP and Repo within the same cell was highlighted in (C) (insert, image represents a Z-projection of 5-12 slices). GFP (green), Repo (red), and Elav (blue). Scale bar ¼ 50 mm. (Krnjevic 1970;Olney and Ho 1970;Fonnum 1984;Choi 1988;Nicholls 1993). Indeed, extracellular glutamate levels are tightly regulated for precise control of neurotransmission at glutamatergic synapses, and to prevent neuronal cell death from excitotoxicity (Danbolt 2001). In Drosophila, this process is regulated by a group of cells that express Eaat1, the only Drosophila high-affinity glutamate transporter (Besson et al. 1999). Although a previous study in adult flies has shown that RNAi-mediated knocking down of Eaat1 in glial cells increased sensitivity to oxidative stress, enhanced degeneration of the brain neuropil and decreased lifespan (Rival et al. 2004), the importance of this group of glial cells in neuronal development and function still remains largely unknown.

Eaat1
First, we determined the developmental expression pattern of Eaat1 by crossing Eaat1-Gal4 with UAS-GFP. As shown in Figure 1, the Eaat1-Gal4 expressing cells were detected throughout all developmental stages from first instar larvae to adulthood. The Eaat1-Gal4 positive cells are a subgroup of glial cell that co-express the glial marker Repo ( Figure 1C). Then, to further determine the role of Eaat1-positive glial cells in development and survival of D.melanogaster, we ablated these cells by specific expression of an apoptotic gene reaper (rpr) in the Eaat1-positive glial cells through crossing the virgin female Eaat1-GAL4 with male UAS-rpr carrying the UAS-rpr transgene on the 2nd or the X chromosome. We found that depletion of this group of cells in both males and females (in the progeny of Eaat1-Gal4 female crossed with male UAS-rpr carrying transgene on the 2nd chromosome) terminated D.melanogaster development at embryonic stage under room air condition, as no pupae and adult flies were obtained. To further confirm this phenotype, we crossed the Eaat1-Gal4 female with male UAS-rpr carrying the transgene on the X chromosome. With this cross, the rpr transgene was only expressed in the female progeny. As expected, only male progeny survived (Figure 2), demonstrating that these glial cells are essential in organismal development and survival.
Neurons and glia are two major cell types in the nervous system of both vertebrate and invertebrate animals. Glial cells are critical for the development and function of neurons by providing neurons with survival and axonal guidance cues, electric shield, and acting as macrophages to remove injured/dead cells (Yildirim et al. 2019;Bittern et al. 2020;Hilu-Dadia and Kurant 2020). Previous studies have shown that Notch signaling plays an important and evolutionarily conserved role in regulating glial differentiation and proliferation during development or following neuronal injuries (Jacobs 2000;Hidalgo and Logan 2017;Kato et al. 2018;Bahrampour and Thor 2020). Our previous study has shown that increasing Notch activity specifically in this set of glial cells dramatically enhanced hypoxia survival, implying that these glial cells also play an important role in regulating organismal development and survival under stress conditions , which, at least in part, through rebalancing hypoxia-induced dysregulation and neurotoxicity of glutamate, maintaining the function of neuropil and development of neuronal-glial system (Choi and Rothman 1990;Stacey et al. 2010;MacNamee et al. 2016). In addition, discovery of the property of Notch signaling in regulating cellular responses to hypoxic challenge further broadened the pleiotropic nature of this signaling pathway Arumugam et al. 2018).

Creation and characterization of a dual-UAS/Gal4 system
In order to determine genetic interactions that mediate the function of Notch signaling under hypoxic conditions, we created a dual-UAS/Gal4 system (Supplementary Figure S1). This system contains  an Eaat1-Gal4 transgene that simultaneously drives the expression of a UAS-NICD (the functional domain of Notch receptor) transgene to activate Notch signaling and a UAS-RNAi transgene to target and knockdown the expression of the specific candidate gene of choice. In order to test the efficiency of this system, we crossed the EN-line with a UAS-LacZ transgene or a background Drosophila strain for the transgenic lines (w 1118 or yw). As shown in Figure 3, the crosses containing Eaat1-Gal4, UAS-NICD, and UAS-LacZ (one copy of each transgene) showed a hypoxia tolerance similar to those containing only Eaat1-Gal4 and UAS-NICD (60%-80%), but significantly higher than those of controls (<25%) (P < 0.01), demonstrating that the presence of an additional UAS (Gal4 upstream activating sequence) did not significantly reduce the efficiency of Gal4-driven expression of UAS-NICD transgene, nor did it significantly affect the phenotype that is induced by NICD overexpression. Hence, the dosage of Eaat1-Gal4 is sufficient to simultaneously drive the expression of two UAS-transgenes.

Evolutionarily conserved mechanisms underlying Notch-conferred hypoxia tolerance
It has been shown that combinatorial and context-dependent interactions between cellular signaling pathways regulate a wide and diverse range of biological processes for development, homeostasis, and disease. Previous studies have demonstrated that cellular context-specific integration of Notch signaling with other genes and pathways are essential for mediating the action of Notch in various developmental and pathological processes (Hurlbut et al. 2007;Bray and Bernard 2010;Borggrefe and Liefke 2012). However, the specific mechanisms by which Notch regulates hypoxia survival still largely remain uncharacterized.
Notch activation-conferred hypoxia tolerance, reducing survival rate from 86.1% to 34.7% in hypoxia (P < 0.01) ( Figure 4G). Furthermore, we found that knocking down bnl, croc and Mkk4 also reduced Notch-conferred hypoxia tolerance (Figure 4, A-C showed hypoxia survival rates that were still significantly higher than those of the controls (Figure 4, D-F). The role of bnl, croc and Mkk4 in Drosophila neuronal development has been reported in various studies. Branchless (bnl) is a fibroblast growth factor (FGF) homolog in Drosophila that is an important regulator of tracheal and neuronal morphogenesis (Barrett et al. 2008;Muha and Muller 2013;Du et al. 2017). A neuronal expression of bnl has been detected in embryo and larval brain, which is essential for the induction of cell migration and neuronal connection (Du et al. 2017). Our results indicate that bnl is also expressed in the Eaat1positive glial cells in addition to neurons. This pattern of expression may be important in neuronal development through regulating cell migration and axon outgrowth to create functional neuronal networks. Furthermore, the expression of bnl is highly inducible by hypoxia, and such expression of bnl is important to trigger the growth of the tracheal system to improve O 2 delivery (Jarecki et al. 1999;Centanin et al. 2008). Becauseinteractions between FGF and Notch signaling has been observed in Drosophila and mammals in particular in the neuronal system (Bartlett et al. 1998;James et al. 2004;Yoon et al. 2004;Ghabrial and Krasnow 2006;Zhou and Armstrong 2007;Fujita et al. 2011;Voelkel et al. 2014), the current results suggest that bnl produced by the Eaat1-positive glial cells is essential for Notch-activation-conferred survival under severe hypoxic conditions, at least in part, by maintaining axon growth and integrity of proper neuronal-glial connectivity during development ( Figure 6).
The crocodile (croc) gene encodes a member of the forkhead transcription factor family in Drosophila. Genes encoding this family of transcription factors are widely conserved during evolution ranging from yeast to humans (Hannenhalli and Kaestner 2009). In Drosophila, the expression of croc is mainly detected in embryo, third instar larvae and adult female flies (Lee and Frasch 2004;Lopez et al. 2017). At the embryonic stage, it is expressed in the head anlagen of the blastoderm and controls the establishment of head structures (Hacker et al. 1995). Although signaling mechanisms regulating the activity of croc are yet uncharacterized, our results indicate that the Notch/croc interaction within the Eaat1-positive glial cells is essential for organismal survival under hypoxic condition, possibly through maintaining the viability and function of these Eaat1-positive glial cells and, thus, neuronal development ( Figure 6).
A number of studies have shown that Mkk4 regulates the activity of JNK pathway and neuronal development in Drosophila (Geuking et al. 2009;Rallis et al. 2010).The interactions between Notch and JNK signaling have been reported in Drosophila and mammals during development or under disease conditions. Such interactions may determine or fine-tune the final output of Notch signaling in a cell context-dependent fashion (Zecchini et al. 1999;Kim et al. 2005Kim et al. , 2010Curry et al. 2006;Ho et al. 2015). It was very interesting to find that cell-specific and ubiquitous manipulation of Mkk4 exhibited opposite phenotypes in terms of hypoxia tolerance (i.e., knocking down Mkk4 specifically in Eaat1-positive glial cells led to lethality, but ubiquitous knocking down this gene enhanced hypoxia survival) ( Figures 4C  and 5), implying a cell context-dependent role of JNK signaling in hypoxia response. Indeed, it has been shown in tumor cells that hypoxia-induced JNK activation increases resistance to chemotherapeutic treatment (Comerford et al. 2004), but, in the pulmonary arteries, such activation leads to structural remodeling (Jin et al. 2000). Furthermore, we found that ubiquitous knocking down of several genes modulating the activity of JNK signaling can enhance hypoxia survival (Azad et al. 2012), which suggests a cell context-dependent role of JNK signaling on hypoxia response in the organism. Therefore, we hypothesize that the interaction between Mkk4 and Notch may regulate survival of glial cells under hypoxic conditions, and, hence, hypoxia tolerance in Drosophila ( Figure 6).
In addition, we performed an experiment with ubiquitous knocking down of these candidate genes using the da-Gal4 driver to evaluate their role in Drosophila development and hypoxia tolerance ( Figure 5). We found that the progeny of [da-Gal4] Â [UAS-RNAibnl] or [da-Gal4] Â [UAS-RNAiEgfr] crosses were lethal under both normoxia and hypoxia conditions. In contrast, flies with da-Gal4>UAS-RNAigrn, da-Gal4>UAS-RNAiinv or da-Gal4>UAS-RNAiMkk4 exhibited significantly enhanced tolerance to hypoxia with > 2-fold eclosion rates as compared to those of the controls under hypoxia (p < 0.01). The result of Mkk4 KD is somewhat surprising: unlike specific knocking down in the Eaat1-positive glial cells (Eaat1-Gal4>UAS-RNAiMkk4) ( Figure 4C), ubiquitous knockdown of Mkk4 significantly enhanced hypoxia tolerance in Drosophila, suggesting a developmental and/or cell type specific role of Mkk4 in regulating organismal response to hypoxia.

Conclusions and perspectives
In conclusion, we evaluated the interactions between Notch signaling and a group of evolutionarily conserved genes in Eaat1-** Figure 5 Effect of ubiquitous knocking down of candidate genes on hypoxia tolerance. Specific UAS-RNAi line targeting bnl, croc, Egfr, grn, h, inv, or Mkk4 was crossed with the da-Gal4 driver to ubiquitously knock down the expression of the targeted genes individually in D.melanogaster. The eclosion rate of the progeny under 5% O 2 hypoxic condition was measured. A significant enhancement of hypoxia survival was detected in the flies with specific knocking down of grn, inv, and Mkk4. In contrast, the crosses with croc and h knocking down showed a similar survival rate with the controls. And ubiquitous knocking down of bnl or Egfr was lethal. Bars represent means 6 SD (n ¼ 3 to 6 vials) for each cross. Ordinary one-way ANOVA [F (8,48) ¼ 31.93, P < 0.0001]with Turkey's multiple comparison tests (**P < 0.01).
positive glial cells and the role of such interactions in regulating hypoxia survival in Drosophila. The synergy of signaling between Notch and bnl, Notch and croc and between Notch and Mkk4 were found to be essential for Notch activation-conferred hypoxia tolerance. We hypothesize that the synergy between these functionally distinct signaling mechanisms regulates cell survival and neuronal development under hypoxic conditions in the Drosophila central nervous system, and this is also sufficient for the survival of the organism as a whole ( Figure 6). Furthermore, we provided evidence indicating that bnl, croc, and Mkk4 are expressed in the Eaat1-positive glial cells. These conserved mechanisms have high potential to be translated into humans for developing strategies to treat hypoxia-induced injuries in the central nervous system.

Data availability
D.melanogaster strains are available upon request. The authors affirm that all data necessary for confirming the conclusions of the article are present within the article, figures, and tables.
Supplementary material is available at G3 online.