Introduction

To cope with oxygen shortage, organisms display evolutionarily conserved responses, most of them dependent on the Hypoxia-Inducible Factor (HIF)1,2. HIF is a heterodimer of basic-Helix-Loop-Helix-PAS (bHLH-PAS) proteins, whose β-subunit is constitutive, and its α-subunit is oxygen-regulated3,4,5. At normal oxygen levels, HIF-α is rapidly degraded at the 26 S proteasome3,4,5, while in hypoxia it stabilizes, enters the nucleus, and after dimerizing with HIF-β, binds hypoxia response elements (HRE) at the regulatory regions of HIF target genes. HIF induces the transcription of a wide array of transcripts that mediate adaptation to hypoxia, including genes involved in angiogenesis, erythropoiesis, vasodilation, and metabolism5,6. The Drosophila orthologues of HIF-α and HIF-β are respectively the bHLH-PAS proteins Sima7 and Tango8, which mediate transcriptional responses to hypoxia in Drosophila that are homologous to those of mammals9,10,11.

Protein synthesis is one of the most energy consuming processes in the cell, and translation rates are predominantly regulated at the level of the initiation step, when ribosomes are recruited to the 5′ CAP structure of mRNAs12,13. 4E-binding proteins (4E-BPs) are translational inhibitors that bind to the Eukaryotic Translation Initiation Factor 4E (eIF-4E), thereby preventing association of eIF-4E with the initiation factor eIF-4G, which is in turn essential for initiation of CAP-dependent translation12,14,15,16. 4E-BPs are inhibited by mechanistic Target of Rapamycin Complex 1 (mTORC1)-dependent phosphorylation, thereby allowing the formation of the CAP-dependent translation initiation complex12,14,15. Conversely, dephosphorylation of 4E-BPs provokes a decrease in global protein synthesis, and activation of alternative initiation mechanisms, such as internal ribosome entry sites (IRES)-dependent translation initiation17.

4E-BPs participate in many cellular processes, including cell growth18,19 and cell survival20, and they have been reported to regulate metabolism21, immune responses22,23, senescence and tumorigenesis24, as well as synaptic structure establishment25, among other processes. One single 4e-bp orthologous gene occurs in Drosophila, and has been named Thor26. Thor participates in developmental processes, as well as in adaptation to different types of stress, including nutritional restriction27 and oxidative stress28.

Three previous system-wide transcriptome analyses detected Thor as one of the genes systematically activated under hypoxia in flies29,30,31. In the current study, we have analyzed the function and regulation of Thor in fly adaptation to hypoxia. We show that Thor is transcriptionally induced in a HIF/Sima-dependent manner; it is required for survival at low oxygen levels, and its loss of function causes mitochondrial alterations and accumulation of ROS, accounting for premature lethality of Thor mutants in hypoxia.

Results

Thor expression is induced in hypoxia

As a first step, we analyzed the expression of Thor under normoxic and hypoxic conditions in different developmental stages. We initially utilized a Thor-lacZ enhancer-trap, which displayed strong β-galactosidase induction in hypoxic embryos, as well as in larval and adult tissues from individuals exposed to 5 h of hypoxia (Fig. 1a and Suppl. Figure 1). Furthermore, Thor-lacZ induction was also observed at longer time periods in midgut cells of adult flies (Suppl. Figure 2), suggesting that Thor is induced in various tissues throughout development in different hypoxic conditions.

Fig. 1
figure 1

Thor transcriptional induction during hypoxia is dependent on Sima and Foxo. (a) X-Gal staining was assessed in normoxia (Nx, 21% O2) and hypoxia (Hx) in the Thor-enhancer-trap fly line (Thor-lacZ) in germ band-extended embryos, 3rd instar larval fat body, and adult indirect flight muscle (3% O2, 5 h for embryos and larvae; 4% O2 5 h for adults). (b) Control (w1118), sima homozygous mutant (simaKG07607) or foxo mutant (foxo21/25) embryos were maintained in normoxia (Nx) or exposed to 3% oxygen for 5 h (Hx), and Thor mRNA levels were assessed by qRT-PCR using RPL29 as an internal control. Data are shown as the mean + standard deviation of three independent experiments with three replicates each. Two-way ANOVA, *p < 0.05; **p < 0.001; n = 3. (c) Thor-enhancer-trap activity was determined in Thor-lacZ embryos overexpressing Sima with an en-Gal4 driver (en-Gal4; UAS-sima) in comparison to control (en-Gal4) embryos.

Full size image

Induction of the enhancer-trap in hypoxia was paralleled by hypoxic induction of Thor mRNA in embryos (Fig. 1b), third instar larvae (Suppl. Figure 3a) and adult flies (Suppl. Figure 3b). This induction depended on the transcription factor Foxo, as previously reported32,33,34,35, and also on the HIF-α orthologue Sima (Fig. 1b; Suppl. Figure 3). Sima protein levels significantly increased in hypoxia in embryos, larvae10,36, as well as in adult flies (Suppl. Figure 4 and Suppl. Figure 5). Moreover, overexpression of Sima was sufficient to induce expression of Thor enhancer-trap activity in normoxia (Fig. 1c), confirming that HIF/Sima controls Thor transcription.

As a next step, we asked whether transcriptional control of Thor by HIF/Sima is direct, so we searched for functional Hypoxia Response Elements (HREs) at the 5’ upstream region of the Thor gene. Four putative HREs bearing the ACGTG consensus sequence (reviewed in37,38,39) occur within 2 kilobases upstream of Thor (Fig. 2a). To analyze if the putative HREs are functional, we cloned fragments of the regulatory region upstream of a luciferase reporter, and transfected these constructs in Drosophila S2 cells along with a Sima expression vector (pAct-sima), or β-galactosidase (pAct-lacZ) as a control. With a fragment of the regulatory region encompassing nucleotides - 1903 to + 52 with respect to the transcription initiation site (TIS), which includes all four presumptive HREs (Fig. 2a), luciferase activity was robustly induced upon overexpression of Sima (Fig. 2b), and this induction depended entirely on the most distal HRE localized at position - 1610 to -1606 (Fig. 2b, c). Consistent with the above results, the complete regulatory fragment (-1903 to + 52) promoted induction of luciferase activity in hypoxia, which was blunted when the distal HRE was mutagenized (Fig. 2d). These results suggest that HIF/Sima exerts direct control of Thor transcription via a functional HRE that occurs 1610 base pairs upstream to the TIS of the Thor gene.

Fig. 2
figure 2

Thor gene is regulated by at least one Hypoxia Response Element. (a) Scheme of the 5’ upstream region of the Thor gene. The sequence and position (relative to the transcription initiation site) of four putative Hypoxia Response Elements (HREs) are indicated. (b) Functional analysis of the putative HREs. Six different reporter constructs of the Thor 5’upstream region were cloned upstream of the Firefly luciferase ORF, and transiently cotransfected in Drosophila S2 cells, along with vectors expressing Sima (pAct-sima), or β-galactosidase (pAct-lacZ) as a control, as well as a HIF-insensitive vector expressing Renilla luciferase (pAct-RenLuc). Fragment 1 (-1903 + 52) includes all four putative HREs; Fragment 2 (-618 + 52) includes only the putative HRE 1; Fragment 3 (-1055 + 52) includes putative HREs 1 and 2; Fragments 4, 5 and 6 include all four putative HREs with point mutations in HRE 3 (Fragment 4), HRE 4 (Fragment 5) or in both HRE 3 and HRE 4 (Fragment 6). The sequences of wild type and mutagenized HREs are shown below Fragment 6. Firefly luciferase activity was normalized to Renilla luciferase activity. Data are shown as change relative to the luciferase activity in cells cotransfected with the Fragment 1 reporter and pAct-lacZ. (c) Sequence of the functional HRE 4 of the Thor 5’ upstream region, encompassing nucleotides - 1610 to -1606 respect to the transcription initiation site, and adjacent nucleotides. (d) The HRE 4 induced hypoxia-dependent expression of the reporter: S2 cells cotransfected with the luciferase reporters based on Fragment 1 or Fragment 5, along with pAct-RenLuc were incubated for 24 h at 21% O2 (Nx) or 1% O2 (Hx). Firefly luciferase activity was normalized to Renilla luciferase activity. Data are shown as fold change over the luciferase activity observed in cells transfected with the Fragment 1 reporter and maintained in Nx. In all cases, data are shown as the mean + standard deviation of three replicates; each graph is representative of three independent experiments. Two-way ANOVA, *p < 0.001, n = 3.

Full size image

Thor is required for Drosophila adaptation to hypoxia

Next, we analyzed the requirement of the Thor gene for organismal adaptation to hypoxia. While homozygous Thor226,40 mutant adult female flies showed normal viability in normal oxygen conditions27, mutants displayed increased lethality in hypoxia (4% O2) in comparison to controls (Fig. 3a). Consistent with this, ubiquitous expression of a Thor RNAi but not of a white RNAi provoked substantial lethality under hypoxia (Fig. 3b). Thus, Thor is necessary for adaptation of flies to hypoxia.

Fig. 3
figure 3

Survival of adult flies in hypoxia is reduced upon Thor loss of function. (a,b) Survival of adult females was assessed in hypoxia (4% O2) at 25 °C comparing Thor2 homozygous mutants and control (yw) flies (a); or Act-Gal4, UAS-ThorRNAi and control (Act-Gal4, whiteRNAi) flies (b). Log-rank (Mantel-Cox) Test, ****p < 0.0001, n = 12 vials of each group containing 10 flies each. Data are represented as Mean +/- SD, and are representative of three independent experiments.

Full size image

4E-BPs are mediators of the effect of mTORC1 on the stimulation of mitochondrial biogenesis and function41,42. Therefore, we analyzed whether Thor mutant flies display mitochondrial alterations that may account for early lethality in hypoxia. Utilizing a transgenic line with mitochondria labelled with the enhanced yellow fluorescent protein (eYFP)43, we observed that Thor homozygous mutant flies displayed enlarged mitochondria as compared to control flies in normoxia (Fig. 4a).

Fig. 4
figure 4

Thor2 flies display mitochondrial alterations. (a) Left upper panel: Confocal image of an indirect flight muscle of a control female transgenic for sqh-eYFP-Mito (green), stained with Alexa546-conjugated phalloidin (red) to visualize F-actin. The yellow dashed line in the inset marks the limits of an individual mitochondrion. Lower images show morphology of mitochondria of control and Thor2 flies at 21% O2 (Nx) or exposed to 4% O2 for 3 days (Hx). Right: Quantification of mitochondrial size. Two-way ANOVA, *p < 0.05; n = 90 mitochondria from 10 flies in each group. Data are representative of three independent experiments. Scale bars: 5 μm. marf (b) and opa1-like (c) transcript levels relative to tubulin were determined by qRT-PCR in control (yw) and Thor2 homozygous female flies maintained at 21% O2 (Nx) or exposed to 4% O2 for 3 days (Hx). Two-way ANOVA, interaction n.s., genotype effect *p < 0.05; n = 4 pools of 8 flies for each group. (d) Mitochondrial DNA copy number was assessed by qPCR (see materials and methods) in relation to the GAPDH nuclear gene in control (yw) and Thor2 homozygous female flies exposed for 3 days to 4% O2 (Hx) or maintained at 21% O2 (Nx). Two-way ANOVA, *p < 0.05; n = 4 pools of 10 flies per group. In all cases, data are presented as Mean + SD.

Full size image

Consistent with this observation, Thor mutants displayed elevated expression of the mitochondrial fusion genes marf and opa1-like, in comparison to wild type flies (Fig. 4b, c). The mitochondrial DNA content, which is usually utilized to estimate the total mitochondrial mass of the cells44, was not statistically different between genotypes in normoxia, suggesting that mitochondrial biogenesis or destruction were not affected in the mutant. Interestingly, mitochondrial DNA increased in Thor mutants but not in control flies exposed to hypoxia (Fig. 4d). Taken together, this set of results suggests that Thor contributes to control mitochondrial function in normoxia and hypoxia.

It has been shown that Thor is required for survival under conditions of oxidative stress27, and that hypoxia tends to provoke accumulation of ROS in the cell45,46. Since Thor2 mutants displayed mitochondrial alterations (Fig. 4) that could potentially lead to ROS accumulation, we explored if augmented ROS levels may account for impaired viability of the mutants in hypoxia. Interestingly, expression of the ROS reporter gstD1-GFP47,48 was not induced in control flies exposed to hypoxia, whereas Thor2 mutant flies exhibited strong hypoxic induction of the reporter (Fig. 5a and Suppl. Figure 6). Furthermore, while hydrogen peroxide did not accumulate in wild type flies exposed to hypoxia, its levels increased by about 50% in Thor2 mutants (Fig. 5b). These results suggest that impaired adaptation of Thor mutant flies to hypoxia may stem from excessive accumulation of ROS. Consistent with this notion, overexpression of the ROS detoxifying enzymes Catalase (Cat) or Superoxide dismutase 2 (Sod2) led to significant reversion of lethality of Thor mutants in hypoxia (Fig. 5c, d).

Fig. 5
figure 5

Thor2 flies accumulate ROS in hypoxia. (a) Control (yw) and Thor2 homozygous female flies expressing the ROS reporter gstD1-GFP48,49 were exposed to 4% O2 (Hx) or maintained at 21% O2 (Nx) for 3 days, and homogenates were prepared for western blot analysis against GFP or tubulin. (b) Hydrogen peroxide levels were determined in control (yw) and Thor2 female flies raised in 10% yeast at 29 °C in normoxia (Nx) or after 3 days of 4% oxygen exposure (Hx). Two-way ANOVA, ***p < 0.001, n = 4 pools of 6 flies per group. Data are representative of three independent experiments. (c) Survival curves of Thor2 homozygous female flies ubiquitously overexpressing the antioxidant enzyme Catalase (Cat) (Thor2; Act5c-Gal4, UAS-cat) or GFP as a control (Thor2; Act5c-Gal4, UAS-eGFP), as well as control transgenic flies expressing only the Gal4 driver (yw; Act5c-Gal4). Flies were exposed to continuous hypoxia (4% O2) for 12 days. Log-rank (Mantel-Cox) Test, **p < 0.01, n = 10 vials of each group containing 10 flies each. Data are representative of two independent experiments. (d) Survival curves of Thor2 homozygous female flies ubiquitously overexpressing the antioxidant enzyme Superoxide dismutase 2 (Sod2) (Thor2; Act5c-Gal4, UAS-sod2) or eGFP as a control (Thor2; Act5c-Gal4, UAS-eGFP), as well as control transgenic flies expressing the Gal4 driver only (yw; Act5c-Gal4). All flies were exposed to hypoxia (4% O2) during 12 days. Log-rank (Mantel-Cox) Test, ****p < 0.0001, n = 10 vials of each group containing 10 flies each. Data are representative of two independent experiments. In all cases, data are presented as Mean +/- SD.

Full size image

Altogether, our results suggest that Thor is required for adaptation of flies to hypoxia, by contributing to shape the mitochondrial behavior, preventing excessive ROS accumulation when oxygen is scarce (Fig. 6).

Fig. 6
figure 6

Model indicating the requirement of Thor for fly adaptation to hypoxia. In hypoxia, Thor is transcriptionally induced in a Sima- and Foxo-dependent manner. Thor is required to prevent accumulation of ROS in hypoxia, by contributing to regulate mitochondria. This Thor function is crucial for fly survival in hypoxia. Illustration created with BioRender.com.

Full size image

Discussion

In this study, we have described a role of the translation inhibitor ThorThor in fly adaptation to hypoxia. We have shown that the Thor gene is transcriptionally induced in multiple Drosophila tissues when they are exposed to hypoxia, and that this induction largely depends on HIF. It has been shown that Foxo, an evolutionary conserved Forkhead Box-O transcription factor, is involved in hypoxic transcriptional induction of multiple genes including Thor in larvae and adult flies32. Our findings complement that study by showing that transcriptional induction of Thor in hypoxia results from the concerted activity of Foxo and HIF.

It has been previously shown that Thor is required for strengthening host immune responses26,50, adaptation to temperature variations51, and survival under conditions of nutrient limitation or oxidative stress27,28,52,53,54,55. Our results extend the above findings, by showing that Thor is also essential for survival under hypoxic conditions. Since Thor is required for setting mitochondrial parameters, thereby regulating ROS production27,28,52,56, we hypothesized that excessive ROS accumulation could be the cause of premature death of Thor mutant flies in hypoxia. We found that indeed, increased levels of ROS accumulated in hypoxic Thor mutants, and particularly high levels of hydrogen peroxide where observed. Overexpression of cat or sod2 led to significant reduction of lethality of Thor mutant flies in hypoxia, confirming that Thor lethality is likely due to augmented ROS levels.

Mitochondria are dynamic organelles in which the frequencies of fission and fusion are balanced to maintain mitochondrial size and function57,58. While hypoxia induces mitochondrial fragmentation in various biological settings, in other situations mitochondrial fusion occurs as an adaptive response to low oxygen tension59,60,61. Our observations in the Drosophila indirect flight muscle support the latter scenario, where mitochondrial size increased in hypoxia, although this was not accompanied by an induction of marf or opa1-like expression. In this case, probably other factors may explain this increment of size, as perhaps other mitochondrial fusion genes or repression of mitochondrial fission genes. Noteworthy, Thor mutants displayed larger-than-normal mitochondria, as well as increased expression of mitochondrial fusion genes already in normoxia, while in hypoxia mitochondrial fusion was not further induced. Regarding mitochondrial total mass, both genotypes were similar in normoxia, while in hypoxia mitochondrial mass of mutant flies was significantly higher than that of controls. Previous data from various experimental systems support a role of 4E-BP/Thor in the regulation of mitochondrial parameters: In the Drosophila indirect flight muscle, overexpression of Thor has been shown to compensate the excessive mitochondrial enlargement that occurs in PINK1 mutant flies62. In mammalian MEFs, 4E-BP was shown to regulate mitochondrial respiration, while its loss of function results in over-accumulation of mitochondrial respiratory complexes42, and conditions that led to increased levels of 4E-BP1/2 provoked reduction of mitochondrial mass41. Taken together, our characterization of mitochondria in Drosophila indirect flight muscles, along with available data from diverse experimental systems, support the notion that Thor plays a central role in the regulation of mitochondria both in normoxia and hypoxia.

Hypoxia induces accumulation of ROS and changes in mitochondrial structure, function and dynamics63,64,65. We have found in our study that in hypoxic conditions, accumulation of ROS, and particularly of hydrogen peroxide, occurs in Thor mutants, leading to increased lethality of the mutants in hypoxia. Teleman et al. have proposed that Thor is a “metabolic brake” that operates under nutritional deprivation to reduce energy consumption, allowing survival in such stress conditions27. We wish to propose that in addition, Thor functions as a “redox buffer” required in hypoxia for maintaining mitochondrial parameters and preventing accumulation of ROS, which is essential for survival in such condition (Fig. 6).

Materials and methods

Fly lines and stocks

Flies were raised at 25 °C in standard cornmeal/agar medium (10 g/l agar; VanRossum, MATPRI0153), 20 g/l or 100 g/l baking yeast, 66.5 g/l cornmeal, 40 g/l sucrose (Anedra, AN00711809), 4.5 ml/l propionic acid (Cicarelli, 1085110), and 0.18 g/l Nipagin (VanRossum, MATPRI0350). Larvae of the desired genotype were sorted 24 h after egg laying. Generally, 30–40 larvae were grown at 25 °C from first to early third instar in plates with enriched medium containing 3.4% baking yeast and 6% sucrose (Anedra, AN00711809).

The following Drosophila melanogaster lines were obtained from the Bloomington Stock Center (http://flystocks.bio.indiana.edu): w1118 (BL3605), yw (BL 64401), Thor2 (BL 9147), Thor-lacZ (BL 9558), act5C-Gal4 (BL 4414), en-Gal4 (BL 30564), UAS-whiteRNAi (BL 33613), UAS-ThorRNAi (BL36667), UAS-sima (BL 9582), UAS-GFP (BL 5430), UAS-sod2 (BL 24494), UAS-cat (BL24621) and sqh-eYFP-Mito (BL 7194). The simaKG07607 mutant, and the gstD1-GFP transgenic reporter fly have been previously described48,66.

Hypoxia treatments

Hypoxia was performed using a Forma Scientific 3131 (Thermo Scientific) incubator or a Ruskinn Invivo2 working station (Ruskinn Technology Ltd) by regulating the proportions of oxygen and nitrogen concentrations according to manufacturer´s protocol.

Survival experiments

Female mated flies aged between 3 and 5 days were divided into groups of 10 individuals and placed in vials containing culture medium. The vials were then introduced into the Ruskinn Invivo2 working station with an oxygen concentration of 4% at 25 °C. Fly survival was assessed daily for a duration of 15 days.

Tissue dissection

Tissues from third-instar larvae or adults were dissected in phosphate-buffered saline (PBS; 8 g/l NaCl, 0.2 g/l KCl, 1.44 g/l Na2HPO4, 0.24 g/l KH2PO4, pH 7.4) and fixed in 4% methanol-free formaldehyde (Polysciences, 18814-10) for 2 h at room temperature. After washing in PBS-0.1% Triton X-100 (Sigma, T-6878), samples were mounted in 40% glycerol for direct visualization.

RNA and DNA purification, and real-time PCR

Total RNA was isolated using the QUICK-ZOL reagent (Kalium Technologies). Genomic DNA was removed from RNA samples using the Ambion’s DNA free kit. RNAs (400 ng) were reverse-transcribed using the Superscript III First-strand synthesis system (Invitrogen) and oligo-dT as a primer. The resulting cDNA was used for real-time PCR (Stratagene MX300 sp), using Taq DNA polymerase Pegasus (PB-L Productos Bio-Logicos) and SYBRGreen and ROX (Invitogen) as fluorescent dyes. Samples were normalized using tubulin primers. Three independent biological samples were analyzed in each experiment. Primer sequences were as follows: Thor fw: GCTAAGATGTCCGCTTCACC; Thor rv: GAAAGCCCGCTCGTAGATAA; Marf fw: TCTACGGCGAGATCGAGGAG; Marf rev: GAGCTCTTGCCGTTGGAGGT; Opa1-like fw: AACGGTGGAGCCAGTTCTCG; Opa1-like rev: TGATCTCCGTCTGCAGCGTC; RPL29 fw: GAACAAGAAGGCCCATCGTA; RPL29 rv: AGTAACAGGCTTTGGCTTGC; Tubulin fw: TGTCGCGTGTGAAACACTTC; Tubulin rev: AGCAGGCGTTTCCAATCTG. After denaturation at 95 °C for 10 min, the amplification conditions were 95 °C for 30 s, 60 °C for 1 min, and 72 °C for 1 min, for 40 cycles, followed by 95 °C for 1 min. Validation of the results were performed using a dissociation curve consisted in ramping the temperature of the samples from 60 to 95 °C, each 30 s, while collecting fluorescent data. For mitochondrial DNA copy number comparisons, real-time PCR was carried-out, using total DNA as a template. Total DNA was extracted by using a standard potassium acetate/phenol-chloroform method. Mitochondrial DNA (mtDNA) target (corresponding to position to 1445–1607 of the mtDNA), and nuclear DNA (nDNA) target (GAPDH) sequences were amplified using following primers: mtDNA fw: CGCCTAAACTTCAGCCACTTAATC; mtDNA rv: AATGCTCCAGGATGTCCTAATTCA, nDNA fw GACGAAATCAAGGCTAAGGTCG; nDNA rv: AATGGGTGTCGCTGAAGAAGTC. Mitochondrial DNA copy number was compared between different conditions by amplifying the mitochondrial DNA target sequence, and using GAPDH amplification as a reference for nuclear genome.

Plasmid constructions

For transient transfection experiments in S2 cells, we employed previously characterized vectors: pAct-lacZ, pAct-LucRen, and pAct-sima67. Six different p5’UTR-Firefly reporter constructs were obtained as follows: The [-1903 + 52] promoter was obtained by PCR of genomic DNA from Drosophila flies. The promoters [-1055 + 52] and [-618 + 52] were generated by PCR amplification of [-1903 + 52] promoter fragments. After adenine addition, all PCR products were cloned into the NheI restriction site of pCR 2.1-TOPO (Invitrogen), and subsequently cloned in pGL3 basic luciferase reporter vector (Promega), both previously digested with NheI/KpnI (New England Biolabs).

Following primers were used:

[-1903 + 52] promoter:

Fw(-1903): 5’- CGCTTAGAGAACCGGCTAGG.

Rv(+ 52): 5’- GCTAGCAGCTTGTGTGGCTGCTATGT (underlined: NheI site).

[-1055 + 52] promoter:

Fw(-1055): 5’-TCCACCTCAAAGAACTTGGCA.

Rv(+ 52).

[-618 + 52] promoter:

Fw(-618): 5’-AGCCTACGTCACGATTGCC.

Rv(+ 52).

Point mutations in the candidate HRE sites 3 and 4 were introduced by overlap extension PCR, where HRE consensus sequence ACGTG was replaced by AAGTG 68,63.

For site 4 the following primers were used:

PCR#1: Fw(-1903) and Rv1: TGCCGTGTCATCTAGACATTAGTCCACAGC;

PCR#2: Fw2: GCTGTGGACTAATGTCTAGATGACACGGCA and Rv(+ 52).

For site 3 the following primers were used:

PCR#1: Fw(-1903) and Rv1: GACAAAAGGTGGCACATTGTGTTGATCTGCGG;

PCR#2: Fw2: CCGCAGATCAACACAAGTGGCCACCTTTTGTC and Rv(+ 52).

WT [-1903 + 52] promoter was used as template for the generation of a construct only mutated in site 3, and [-1903 + 52] promoter mutated in site 4 was used as template for generating a construct mutated in both sites 4 and 3.

Cell culture, transfections and luciferase assay

Drosophila Schneider cell line (S2) was cultured in Schneider Drosophila medium (Sigma), supplemented with 10% fetal bovine serum (Invitrogen), Penicillin (50 U/ml, Invitrogen), and Streptomycin (50 ug/ml, Invitrogen) at 25 °C in 25 cm2 or 75 cm2 T-flasks (Greiner). Cells were exposed to hypoxia in a Forma Scientific 3131 (Thermo Scientific) incubator.

For transient transfection experiments, cells were seeded in 24-well plates (Greiner) at a 350,000 cells per well in 350 µl of culture media, and 0.15 µg of total plasmid DNA was transfected employing the Effectene transfection reagent (Qiagen), as follows: 87.5 ng of each promoter-LucFF construct, 52.5 ng of pAct-sima or pAct-lacZ vectors, and 10 ng of pAct-LucRen (Renilla). Each transfection was performed in triplicate. Expression of luciferase from the different promoters-LucFF vectors was recorded 24 h after transfection. Firefly and Renilla luciferase activities were measured by the Dual-Glo Luciferase Assay System (Promega) following manufacturer’s instructions in a Veritas Microplate Luminometer (Turner BioSystems).

X-gal staining

Transgenic embryos, larvae, and adult female flies bearing the lacZ reporter gen driven by the Drosophila Thor promoter (Thor-lacZ) were used.

For X-gal stainings, bleach-dechorionated embryos incubated with heptane for 5 min, or dissected organs from larvae or adults were fixed with 0.5% glutaraldehyde in PBS during 10 min at room temperature, and washed in PT (0.1% Triton X-100 in PBS). Samples were incubated in staining solution (5 mM K4Fe(CN)6, 5 mM K3Fe(CN)6, 0.2% X-gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside) in PT) at 37 °C, and the colorimetric reaction was monitored. Reactions were stopped by three washes with PT, and photographed with a Nomarski Olympus BX-60 microscope equipped with an Olympus DP71 digital camera.

Mitochondrial analysis

For mitochondrial visualization, indirect flight muscles from 3-day-old female adults of the sqh-eYFP-Mito stock were dissected, fixed for 30 min in methanol-free formaldehyde, washed three times for 10 min each, and incubated in Alexa546-conjugated phalloidin (1:250 in PBS-0.1% Triton X-100, ThermoFisher Scientific, A22283) for 30 min to stain actin filaments. Finally, samples were washed in PBS-0.1% Triton X-100 and mounted in 40% glycerol for direct visualization. Images were obtained using a Carl Zeiss LSM 710 confocal microscope with a Plan-Neofluor 40×/1.3 NA objective, and mitochondrial size was determined using the polygon selection tool followed by area measurement with the ImageJ software (NIH). Quantification of mitochondria area was assessed using the ImageJ software with the freehand selection tool followed by area measurement.

Western blots

Western blots were carried out by standard procedures using ECL plus (GE, RPN2232). The primary antibodies used were rabbit anti-Sima-ODD10, rabbit anti-GFP (Invitrogen #A6455; 1/5,000) and mouse anti–α Tubulin (Developmental Studies Hybridoma Bank—DSHB #12G10; 1/10,000). Secondary peroxidase–conjugated antibodies used were donkey anti-rabbit (Jackson ImmunoResearch #111-035-144; 1/5,000) and donkey anti-mouse (Jackson ImmunoResearch #715-035-150; 1/5,000).

Hydrogen peroxide determination

Hydrogen peroxide was measured using the Amplex Red Hydrogen Peroxide Assay kit (Invitrogen) according to manufacturer’s instructions. Briefly, female flies submitted to 4% oxygen for 3 days at 29 °C, were homogenized in 300 µl Krebs Ringer phosphate solution and centrifuged at 5000xg for 10 min at 4 °C. Then, working solution (50 µl; 100 µΜ Amplex Red Reagent and 0.2 U/ml HRP) was added to 50 µl of a 1:10 dilution of the supernatant. After 30 min incubation at room temperature in the dark absorbance at 560 nm was measured in a SmartSpec 3000 spectrophotometer (Bio-Rad Laboratories). Hydrogen peroxide concentration was calculated using a standard curve, and finally normalized to protein levels. Experiments were performed in quadruplicate.

Statistics

Statistical analyses were performed using GraphPad Prism 9.0.0 software. Statistical significance was calculated using two-way analysis of variance (ANOVA) followed by Fisher, Bonferroni, or Tukey’s tests with a 95% confidence interval (p < 0.05). In cases where comparison between two treatments was needed, two-tailed, unpaired Student’s t-test was used instead. For survival curves, Log-rank (Mantel-Cox) test was used. Data were represented as mean +/- SD. Each experiment was independently repeated at least twice.