Author Response: Non-canonical Function of an Hif-1α Splice Variant Contributes to the Sustained Flight of Locusts
Ding,Jie Zhang,Beining Du,Xuanzhao Wang,Hou Li,Siyuan Guo,Bing Chen,Le Kang
DOI: https://doi.org/10.7554/elife.74554.sa2
2022-01-01
Abstract:Article Figures and data Abstract Editor's evaluation Introduction Results Discussion Materials and methods Data availability References Decision letter Author response Article and author information Metrics Abstract The hypoxia inducible factor (Hif) pathway is functionally conserved across metazoans in modulating cellular adaptations to hypoxia. However, the functions of this pathway under aerobic physiological conditions are rarely investigated. Here, we show that Hif-1α2, a locust Hif-1α isoform, does not induce canonical hypoxic responses but functions as a specific regulator of locust flight, which is a completely aerobic physiological process. Two Hif-1α splice variants were identified in locusts, a ubiquitously expressed Hif-1α1 and a muscle-predominantly expressed Hif-1α2. Hif-1α1 that induces typical hypoxic responses upon hypoxia exposure remains inactive during flight. By contrast, the expression of Hif-1α2, which lacks C-terminal transactivation domain, is less sensitive to oxygen tension but induced extensively by flying. Hif-1α2 regulates physiological processes involved in glucose metabolism and antioxidation during flight and sustains flight endurance by maintaining redox homeostasis through upregulating the production of a reactive oxygen species (ROS) quencher, DJ-1. Overall, this study reveals a novel Hif-mediated mechanism underlying prolonged aerobic physiological activity. Editor's evaluation The hypoxia-inducible factor (Hif) pathway is well known for modulating cellular responses to hypoxia. Combining physiological and molecular biology insights along with evolutionary analysis, this study shows that a unique Hif-1a splice variant remains active in normoxia and scavenges flight-induced ROS to protect the muscle during the prolonged flight in locusts. This work is novel and as a study in a non-model system, is particularly noteworthy for its elegance. https://doi.org/10.7554/eLife.74554.sa0 Decision letter Reviews on Sciety eLife's review process Introduction Flight is a key adaptive strategy for many animals but also poses physiological challenges. Animal flight is the most energetically expensive form of locomotion that relies heavily on mitochondrial aerobic activity. The oxygen consumption rates during flight are approximately 30–150 times higher than those at rest (Armstrong and Mordue, 1985; Bartholomew and Casey, 1978; Snelling et al., 2012). However, the high aerobic performance of flight muscles can produce excessive amounts of reactive oxygen species (ROS) and cause oxidative damage to the myocytes (Fisher-Wellman and Bloomer, 2009). The flight activity of Drosophila is associated with reduced life span and increased levels of oxidative damage, but oxidative-induced muscle fatigue is rarely observed in flying animals during long-distance migration (Liechti et al., 2013; Magwere et al., 2006). Flight muscles of a long-distance flying moth use the pentose phosphate pathway (PPP) in a way that appears to reduce oxidative damage caused by flight (Levin et al., 2017). However, under flight conditions, especially during prolonged and continuous flight, the molecular signaling by which flying animals minimize oxidative damage in their muscle systems remains unknown. The hypoxia inducible factor (Hif) pathway is an evolutionarily conserved oxygen sensing pathway that modulates cellular oxygen level and promotes metabolic responses upon hypoxic induction in animals. This pathway controls hundreds of downstream gene expressions and is an important therapeutic target for many hypoxia-associated diseases in human (Kim and Kaelin, 2004; Masoud and Li, 2015; Selak et al., 2005). The key components of this pathway are Hifs, which comprise an oxygen-sensitive α subunit and a stable β subunit. Under hypoxia, stabilized α subunits heterodimerize with β subunits in the nucleus and induce expression of multiple genes (Wenger et al., 2005). In skeletal muscles of mammals, acute exercises lead to reduced myocellular oxygen partial pressure (PO2) and thus stabilize Hif-1α protein, which in turn attenuates ROS production by reducing mitochondrial activity and cellular oxygen consumption (Ameln et al., 2005; Mason et al., 2004). However, flying animals are able to satisfy the high oxygen demand by the myocytes during flight (Komai, 1998; Meir et al., 2019). In particular, owing to invagination of the tracheoles into the cells, flying insects conduct oxygen directly into flight muscle cells and exhibit the highest aerobic activity among animals (Harrison and Roberts, 2000). Therefore, whether or not the Hif pathway is also involved in flight, a constant physical activity that heavily depends on mitochondrial aerobic metabolism, and how the Hif pathway provides protection against oxidative damage and satisfies the extremely high demands of oxygen consumption are unknown. In mammals, three Hif-α paralogs (Hif-1α, Hif-2α, and Hif-3α) have been identified; they are derived from genome duplication and encoded by three separated genes, namely, HIF1A, EPAS1, and HIF3A (Kaelin and Ratcliffe, 2008). Hif-1α is present in all tissues, whereas Hif-2α and Hif-3α are limited to specific cell types (Bertout et al., 2008). Hif-1α enhances oxygen delivery and regulates cellular metabolic adaptation to hypoxia, while Hif-2α promotes vascular endothelial growth by erythropoietin (Iyer et al., 1998; Rankin et al., 2007; Zhang et al., 2007). Due to alternative splicing and different promoters, Hif-3α produces a large number of mRNA variants. The most studied Hif-3α truncated isoform is an inhibitor of Hif-1α and Hif-2α (Maynard et al., 2005). Invertebrates possess only one Hif-α gene (Hif-1α) (Loenarz et al., 2011). But flybase indicates four alternative transcripts in Drosophila and there is further evidence for splice variants in the transcriptomes of other insect species. In insects, the Hif pathway participates in a series of hypoxia-associated biological processes, including tracheal development, diapause, and larval growth (Centanin et al., 2010; Chen et al., 2021; Valzania et al., 2018). In Glanville Fritillary butterfly (Melitaea cinxia), a genetic variation of Sdhd, which is a regulator of Hif-1α, is associated with altered tracheal volume and flight performance; this association suggests the possible involvement of Hif signaling in flight adaptation (Marden et al., 2013; Marden et al., 2021; Pekny et al., 2018). The migratory locust (Locusta migratoria) is a worldwide agricultural pest capable of long-distance flight (Ma et al., 2012; Wang et al., 2014). Swarming locusts can migrate hundreds of kilometers per day and invade areas covering millions of square kilometers to an extent, with the help of thermal currents and tail winds (Zhang et al., 2019). Moreover, migratory locusts can metabolically adapt to hypoxic environments and make migratory flights on the Tibetan Plateau at altitudes over 3700 m (Ding et al., 2018; Zhang et al., 2013). As a model for animal flight studies, this species has been employed to reveal many aspects and processes pivotal to the understanding of hormonal regulation of energy material mobilization during flight (Van der Horst and Rodenburg, 2010). In this study, two alternatively spliced Hif-1α isoforms, namely, Hif-1α1 and Hif-1α2, were identified in migratory locusts. The stabilization of Hif-1α1 requires hypoxic induction, whereas, owing to the lack of C-terminal transactivation domain (C-TAD), Hif-1α2 is less sensitive to oxygen and remains stable in normoxia. In vivo, Hif-1α1 plays a classical role in mediating cellular responses to hypoxia. However, Hif-1α2 is an essential regulator of prolonged flight in locusts. Mechanistically, Hif-1α2 is upregulated extensively during flight and modulates glucose oxidation and redox homeostasis in flight muscles. Hif-1α2 is able to transactivate the expression of Dj-1, which is an important ROS quencher, and scavenges flight-induced ROS directly. These findings reveal a regulatory mechanism by which hypoxia signaling balances high aerobic metabolism and the risk of overloaded peroxidation under highly aerobic physiological conditions. Results Identification of Hif-1α isoforms in locusts The full-length coding sequences (CDs) of Hif-1α (3309 bp) were cloned from locusts. Similar to any other invertebrates, there is only one Hif-α paralog in locusts (Figure 1A). The predicted protein of locust Hif-1α contained basic helix–loop–helix-PER-ARNT-SIM (bHLH-PAS) domains and two transactivation domains (N-terminal TAD and C-terminal TAD). Sequence alignment analysis revealed that the locust Hif-1α protein shared 51.79%, 51.83%, and 48.3% identity with human Hif-1α (AAC68568.1), Hif-2α (Q99814.3), and Hif-3α (AAH80551.1), respectively. In addition, the locust Hif-1α was conserved at proline and asparagine residues, which are the targets of prolyl hydroxylases (PHDs) and factor-inhibiting Hif (FIH), respectively (Figure 1—figure supplement 1). On the basis of full-length transcriptome, two transcripts of Hif-1α were observed. The full and short length were named as Hif-1α1 and Hif-1α2, respectively. These transcripts were both derived from the alternative 3′ splice site coupled to the alternative polyA site (Figure 1B), and further determined by 3′ RACE (GeneBank access numbers MW349109 for Hif-1α1 and MW349110 for Hif-1α2). The CDs of Hif-1α2 were 201 bp shorter than those of Hif-1α1 and the encoded protein lacked the C-TAD, which is required for the transitivity of Hif-1α (Figure 1C and D). Evolutionary analysis revealed that such Hif-1α splice form also exists in other Orthoptera (Accession no. ON137898 and ON137899 for Deracantha onos), some birds (e.g., XP_025006307.1 for Gallus gallus and XP_013038471.1 for Anser cygnoides domesticus), and human (NP_851397.1). Additionally, the TADs of Hif-α have varied distributions among insects. In incomplete metamorphosis insects and beetles, the Hif-α protein possesses two TADs (N-TAD and C-TAD), but in flies and moths the C-TAD and its inhibitor FIH are completely missing at the genomic level. Therefore, C-TAD-lacking Hif-1α transcripts, with distinct origins, seem to commonly exist in different insect taxa (Figure 1—figure supplement 2). The protein of overexpressed Hif-1α1 was detectable only under hypoxic conditions (1% O2), while the protein of Hif-1α2 was detectable in normoxia and upregulated significantly in hypoxia (Figure 1E). Additionally, the protein of Hif-1α2 was observed in the nucleus of cells (Figure 1—figure supplement 3). In adult locusts, Hif-1α1 mRNA was ubiquitously expressed in all tested tissues, but Hif-1α2 showed the highest expression in flight muscles (Figure 1F). In addition, the mRNA levels of Hif-1α2 were elevated with the maturation of the flight muscles and reached the highest level at the adult stage 12 days post eclosion, while Hif-1α1 exhibited stable expression at all developmental stages (Figure 1G). These results suggest a possible functional division between Hif-1α1 and Hif-1α2 in locusts. Figure 1 with 3 supplements see all Download asset Open asset An oxygen-insensitive Hif-1α isoform is identified in locusts. (A) Phylogenetic tree construction of Hif-α. Amino acid sequences of Hif-α family genes from Homo sapiens, Mus musculus, Xenopus laevis, Danio rerio, Branchiostoma floridae, Limulus polyphemus, Zootermopsis nevadensis, Tribolium castaneum, Manduca sexta, Drosophila melanogaster, and Locusta migratoria were used for analysis. One thousand bootstraps were performed. Only bootstrap values of over 60 are shown. (B) Schematic of Hif-1α and its two alternatively spliced isoform mRNAs, Hif-1α1 and Hif-1α2. (C) Hif-1α1 and Hif-1α2 mRNA expression analyzed by RT-PCR. (D) Schematic of Hif-1α1 and Hif-1α2 proteins. (E) In vitro expression of Hif-1α1 and Hif-1α2 under different oxygen concentrations. As shown in the Western blotting, Hif-1α2 protein is detectable in normoxia and hypoxia (1% O2), while Hif-1α1 is detectable only in hypoxia (n=3 replicates). (F) Tissue distributions of Hif-1α1 and Hif-1α2 (n=3 replicates, 5 locusts/replicate). (G) Expression profiles of Hif-1α1 and Hif-1α2 in thoracic muscle during development (fourth- and fifth-instar nymphs and first-, sixth-, and 12th-day adults; n=4 replicates, 5 locusts/replicate). The values of the columns are shown as mean ± standard error (s.e.m.). One-way ANOVA with Bonfferroni’s test for multiple comparisons, significant differences are denoted by different letters or **p<0.01 and ***p<0.001. Figure 1—source data 1 Raw data for spatiotemporal expression profiles of Hif-1α1 and Hif-1α2. https://cdn.elifesciences.org/articles/74554/elife-74554-fig1-data1-v1.xlsx Download elife-74554-fig1-data1-v1.xlsx Figure 1—source data 2 Raw data for mRNA and protein expression levels of Hif-1α1 and Hif-1α2. https://cdn.elifesciences.org/articles/74554/elife-74554-fig1-data2-v1.docx Download elife-74554-fig1-data2-v1.docx Hif-1α1 regulates typical cellular responses to hypoxia To explore the in vivo functions of Hif-1α1 and Hif-1α2 associated with the hypoxic response, hypoxia treatment combined with RNAi assay was performed. Hif-1α1 and Hif-1α2 were knocked down by double-strand RNA (dsRNA) injection (Figure 2A), and the locusts were subjected to short-term hypoxia (3.5 kPa PO2 for 2 hr). Compared with those of the normoxic control, the hemolymph lactate levels were elevated significantly under hypoxia, and this elevation was attenuated by the knockdown of Hif-1α1. Meanwhile, the activities of pyruvate dehydrogenase (PDH), which are negatively regulated by Hif pathway, were significantly repressed under hypoxia and were normalized by the knockdown of Hif-1α1. By contrast, the knockdown of Hif-1α2 had no effects on hypoxia-induced generation of lactate and repression of PDH activity (Figure 2B). Additionally, the flight muscle mitochondrial DNA copy number in the dsHif-1α1-injected groups, but not in the dsHif-1α2-injected groups, was reduced significantly under prolonged hypoxia (3.5 kPa PO2 for 13 hr) compared with that in dsGFP-injected groups (Figure 2C). Thus, the knockdown of locust Hif-1α1 undermined hypoxic responses and produced cellular damages under hypoxia. Further examination was conducted on the regulatory roles of Hif-1α on PDK and BNIP3, which account for the Hif-mediated PDH activity reduction and mitochondrial biogenesis inhibition. Knockdown of Hif-1α1 and Hif-1α2 had no effect on the mRNA expression of PDK and BNIP3 under normoxic condition (Figure 2—figure supplement 1A). But compared with those of the normoxia control, the expression levels of PDK and BNIP3 were upregulated significantly in response to hypoxic induction, and these hypoxia-induced upregulations were repressed by the knockdown of Hif-1α1 (Figure 2—figure supplement 1B). For further confirmation, the Hif pathway was activated in normoxia by silencing PHD, an inhibitor of Hif-α (Figure 2—figure supplement 1C). The mRNA expression levels of PDK and BNIP3 were significantly elevated with the knockdown of PHD. When Hif-1α1 and Hif-1α2 were further knocked down in PHD silenced locusts, these upregulations were attenuated by the knockdown of Hif-1α1 (Figure 2—figure supplement 1D). However, the knockdown of Hif-1α2 did not exhibit obvious effects on PDK and BNIP3 expressions under hypoxic or PHD silencing conditions (Figure 2—figure supplement 1B and D). Therefore, Hif-1α1, rather than Hif-1α2, modulates the cellular responses to hypoxia in locusts. Figure 2 with 3 supplements see all Download asset Open asset Hif-1α1 regulates systematic hypoxic responses, Hif-1α2 is critical for prolonged flight. (A) Knockdown of Hif-1α1 and Hif-1α2 via injections of dsHif-1α1 and dsHif-1α2, with dsGFP-injected groups as control (n=5 replicates, 3 locusts/replicate). (B) Hemolymph lactate production and pyruvate dehydrogenase (PDH) activity tests upon short-term hypoxic (3.5 kPa PO2 for 2 hr) exposure (n=5 replicates, 3 locusts/replicate). (C) Mitochondrial DNA copy number measurements upon long-term hypoxic (3.5 kPa PO2 for 13 hr) exposure (n=5 replicates, 3 locusts/replicate). (D) Expression levels of Hif-1α1 and Hif-1α2 under flight treatment (n=9 replicates, 1 locust/replicate). (E) Flight performance tests, data for each time point was obtained independently (n≥34). (F) Muscle lactate level, PDH activity, and mitochondrial DNA copy number measurements upon flight treatment (n=5 replicates, 3 locusts/replicate). (A and E) Student’s t-test for pairwise comparisons, with dsGFP-injected groups as control. (B, D, and F) One-way ANOVA with Bonfferroni’s test for multiple comparisons. Values are mean ± s.e.m. Significant differences are denoted by *p<0.05, **p<0.01, and ***p<0.001; n.s. represents no significant difference. Figure 2—source data 1 Raw data for mRNA expression profile of Hif-1α1 and Hif-1α2 under flight treatment. Raw data for lactate content, pyruvate dehydrogenase (PDH) activity, mitochondrial DNA copy number measurement, and flight performance test in the absence of Hif-1α1 and Hif-1α2, respectively. https://cdn.elifesciences.org/articles/74554/elife-74554-fig2-data1-v1.xlsx Download elife-74554-fig2-data1-v1.xlsx Hif-1α2 is indispensable for sustained flight of locusts Hif-1α2 is possibly associated with the flight performance of locusts due to its muscle-specific high expression. Thus, forced flight treatment was conducted on locusts and gene expression profiles were examined. Compared with static control, the transcript levels of Hif-1α2 were significantly upregulated at 30 min of flight (p<0.001) and remained by over twofold at 60 min of flight (p<0.01). By contrast, the transcript levels of Hif-1α1 were not upregulated by flight treatment (Figure 2D). Furthermore, the effects of Hif-1α on the flight performance were investigated using computerized flight mills (Figure 2—figure supplement 2). The results showed that the knockdown of Hif-1α2 extensively impaired the prolonged flight performance of locusts. Compared with those of the control (injected with dsGFP), the flight distance, flight duration, and average flight speed of the Hif-1α2-knockdown locust at 120 and 180 min on flight mills were significantly reduced. By contrast, the knockdown of Hif-1α1 had no obvious effects on flight performance (Figure 2E). To explore the activities of Hif-1α1-mediated hypoxic responses during flight, the mRNA levels of PDK and BNIP3 were tested. Upon flight treatment, no significant upregulations of PDK and BNIP3 mRNA were observed (Figure 2—figure supplement 3A). The knockdown of Hif-1α1 and Hif-1α2 during flight had no effects on the mRNA expressions of PDK and BNIP3 (Figure 2—figure supplement 3B). Moreover, the muscle lactate content, PDH activities, and mitochondrial copy number were measured. Likewise, no changes in muscle lactate levels, PDH activities, and mitochondrial copy number in response to flight or to the knockdown of Hif-1α1 and Hif-1α2 were observed (Figure 2F). Thus, Hif-1α2 is indispensable for sustained flight of locusts, and flight does not trigger typical hypoxic responses, which are modulated by Hif-1α1 in flight muscles. Hif-1α2 regulates glucose metabolic and antioxidative gene expressions in locusts To determine the regulatory mechanisms of Hif-1α2 in flight performance, we conducted transcriptomic analysis. The gene expression levels of the Hif-1α1 and Hif-1α2 knockdown groups were compared with those of the dsGFP-injected group. Under normoxic conditions, the knockdown of Hif-1α1 had a slight effect on the gene expression patterns in flight muscles. No genes had false discovery rate (FDR) lower than 0.05. In the Hif-1α2 knockdown group, 12 downregulated genes (MIOX, GAPDH, PGX, GBE, PGI, ENO, DJ-1, PFK, PGM, PYK, LDH, and PDHX) were identified as the candidate target genes (FDR <0.05 and fold change ≥2; Figure 3A and Figure 3—figure supplement 1). The expression patterns of these genes, except MIOX, were confirmed by quantitative real-time PCR (qRT-PCR, Figure 3B). Then, PHD was silenced to activate the Hif pathway. Meanwhile, Hif-1α2 was knocked down and the expression patterns of the candidate target genes were tested. Except MIOX, all genes were significantly upregulated in the PHD knockdown group, and these upregulations were extensively attenuated by the knockdown of Hif-1α2 (Figure 3B). Herein, the 11 target genes of Hif-1α2 were identified. The proteins encoded by GAPDH, PGX, GBE1, PGI, ENO, PFK, PGM, PYK, and LDH are enzymes involved in glycolysis. The protein encoded by PDHX is a core component of the PDH complex, and the protein encoded by DJ-1 is involved in antioxidation (Figure 3C). The regulatory effects of Hif-1α2 on these processes mainly occur in flight muscles and fat bodies but not in the midgut (Figure 3—figure supplement 2). Thus, under normoxic conditions, the locust Hif-1α2 may play major roles in regulating glucose metabolism and antioxidative defense. Figure 3 with 2 supplements see all Download asset Open asset Hif-1α2 target gene identification in flight muscle. (A) Volcano plot showing a global gene expression pattern of Hif-1α1 and Hif-1α2 knockdown groups compared with that of dsGFP-injected group in normoxia under resting conditions. Downregulated genes are colored in green, and upregulated genes are colored in red. The top 12 differentially expressed genes with false discovery rate (FDR) values lower than 0.05 are noted by their gene names and identified as candidate target genes of Hif-1α2. (B) Quantitative real-time PCR (qRT-PCR) confirmation of the expression patterns of the 12 candidate target genes. Columns are mean ± s.e.m. Significant differences are denoted by different letters (one-way ANOVA with Bonfferroni’s test for multiple comparisons, p<0.05, n=4–5 replicates, 5 locusts/replicate). (C) Two biological processes regulated by Hif-1α2 under normoxic conditions. Figure 3—source data 1 Raw data for quantitative real-time PCR (qRT-PCR) verifications of Hif-1α2 target genes. https://cdn.elifesciences.org/articles/74554/elife-74554-fig3-data1-v1.xlsx Download elife-74554-fig3-data1-v1.xlsx Flight-induced elevation of glucose metabolism is Hif-1α2-dependent To further determine the regulatory effects of Hif-1α on glucose metabolism, the activities of pyruvate kinase (PYK) and phosphofructokinase (PFK) were measured. PYK and PFK are rate-limiting enzymes of glycolysis. Knockdown of Hif-1α2, but not Hif-1α1, extensively repressed the activities of PYK and PFK (Figure 4A). Then, the glycolytic gene expression was examined when the locusts were forced to fly for 60 min after the knockdown of Hif-1α1 and Hif-1α2. The expression levels of GAPDH, PGX, GBE1, PGI, ENO, PFK, PGM, PYK, LDH, and PDHX were tested via qRT-PCR. The results showed that knockdown of Hif-1α2, rather than Hif-1α1, dramatically repressed flight-induced upregulation of these genes (Figure 4B). In addition, the hemolymph trehalose content reduced significantly in post-flight locusts, but knockdown of Hif-1α1 and Hif-1α2 demonstrated no significant effect on hemolymph trehalose level. However, the post-flight locusts exhibited an accumulated muscle glucose content and a reduced pyruvate level in the Hif-1α2 knockdown group (Figure 4C). Thus, the reduction in Hif-1α2 has no obvious effect on hemolymph trehalose uptake but remarkably impairs glucose utilization in flight muscles during flight. Figure 4 Download asset Open asset Hif-1α2 regulates glucose metabolism during flight. (A) Enzyme activities of pyruvate kinase (PYK) and phosphofructokinase (PFK) repressed by knockdown of Hif-1α2 in flight muscles under resting conditions (n=5–7 replicates, 3 locusts/replicate). (B) Flight-induced upregulations of glycolytic genes repressed by knockdown of Hif-1α2 (n=4 replicates, 3 locusts/replicate). Heat map signal indicates log2 fold change values relative to the mean expression within each group. (C) Knockdown of Hif-1α2 demonstrated no effect on hemolymph trehalose content but inhibited glucose utilization and pyruvate generation in flight muscle during flight (n=5 replicates, 3 locusts/replicate). (D) No effect of PYK knockdown (n=5 replicates, 3 locusts/replicate) on long-term flight performance (n≥20). (A and D) Student’s t-test for pairwise comparisons, with dsGFP-injected groups as control. (B and C) One-way ANOVA with Bonfferroni’s test for multiple comparisons. Values are mean ± s.e.m. Significant differences are denoted by *p<0.05, **p<0.01, and ***p<0.001; n.s. represents no significant difference. Figure 4—source data 1 Raw data for enzyme activities, metabolite contents, and Hif-1α2 target gene expression profiles, and raw data for flight performance tests in the absence of pyruvate kinase (PYK). https://cdn.elifesciences.org/articles/74554/elife-74554-fig4-data1-v1.xlsx Download elife-74554-fig4-data1-v1.xlsx To verify the effect of glucose metabolism on prolonged flight, glycolysis was repressed by silencing PYK and the flight performance of the locust was tested. The reduction in PYK could repress glycolysis by attenuating the production of pyruvate. However, when the PYK in locusts was knocked down, no obvious effects were observed on flight distance, flight duration, and average flight speed in 60, 120, and 180 min tests (Figure 4D). Therefore, regulating glucose metabolism may not be the main function of Hif-1α2 in sustaining long-term flight. Hif-1α2 plays an antioxidative role during prolonged flight via Dj-1 In addition to glucose metabolism, the target genes of Hif-1α2 are involved in antioxidative defense. We thus knocked down Hif-1α1 and Hif-1α2 and tested the effects of flight on oxidative stress. The ROS levels of the flight muscle were measured by dihydroethidium (DHE) staining. At rest, the locusts in the Hif-1α1 or Hif-1α2 knockdown and control groups did not exhibit differences in DHE signals. However, in post-flight locusts, the DHE signals in the Hif-1α2 knockdown group were significantly elevated compared with those in the control. By contrast, the knockdown of Hif-1α1 displayed no significant differences (Figure 5A). After flight treatment, a significant elevation in the H2O2 content and a reduction in the ratios of reduced to oxidized glutathione (GSH/GSSG) were observed in the flight muscles of the Hif-1α2 knockdown locusts, but not in the Hif-1α1 knockdown locusts (Figure 5B). Therefore, in the flight muscle of locusts, the reduction in Hif-1α2 promotes flight-induced ROS generation and impairs the ability to recycle GSSG to its reduced active form (GSH) to control oxidative damage. Figure 5 Download asset Open asset Hif-1α2 regulates redox homeostasis during flight. (A) High levels of reactive oxygen species (ROS) in flight muscle in the knockdown of Hif-1α2 locusts during flight. Scale bars: 200 µm. Dihydroethidium (DHE) staining intensity was quantified using ImageJ software (https://imagej.nih.gov/ij, n=5–8 replicates). (B) Promotion of H2O2 production and reduction in the ratios of reduced to oxidized glutathione (GSH/GSSG) by knockdown of Hif-1α2 during flight (n=5–6 replicates, 3 locusts/replicate). One-way ANOVA with Bonfferroni’s test for multiple comparisons. Values are mean ± s.e.m. Significant differences are denoted by *p<0.05 and **p<0.01; n.s. represents no significant difference. Figure 5—source data 1 Raw data for dihydroethidium (DHE) staining intensity, H2O2 content, and reduced to oxidized glutathione (GSH/GSSG) level. https://cdn.elifesciences.org/articles/74554/elife-74554-fig5-data1-v1.xlsx Download elife-74554-fig5-data1-v1.xlsx Next, the Hif downstream signaling involved in oxidative stress regulation during locust flight was observed. The main focus was DJ-1, one of the target genes that was regulated by Hif-1α2 in locusts (Figure 3C). The encoded protein of DJ-1 is an evolutionarily conserved ROS quencher (Figure 6—figure supplement 1A-C). In locusts, DJ-1 displayed the highest expression in the maturated flight muscles, it shared a similar expression pattern with that of Hif-1α2 (Figure 6—figure supplement 1D). In addition, the flight-induced upregulations of DJ-1 were extensively repressed by the knockdown of Hif-1α2 (Figure 6A). At the promoter region of DJ-1, three hypoxia response elements (HREs, 5′-RCGTG-3′) were identified using the MatInspector software (Cartharius et al., 2005). In vitro luciferase reporter assay demonstrated that the overexpression of Hif-1α2 increased the transcriptional activity of the DJ-1 promoter under normoxia (Figure 6B). Electrophoretic mobility shift assays (EMSA) demonstrated that Hif-1α2 could bind to the DNA probe that contained the HRE which was located in the position between –1878 and –1874 bp of DJ-1 promoter (Figure 6C). Thus, DJ-1 could be a direct target of Hif-1α2 in locusts. Figure 6 with 2 supplements see all Download asset Open asset Hif-1α2 regulates redox homeostasis via Dj-1. (A) Hif-1α2-dependent flight-induced upregulations of DJ-1 (n=4 replicates, 3 locusts/replicate). (B) Dual luciferase assay demonstrating the effect of Hif-1α1 and Hif-1α2 overexpression on DJ-1 promoter (n=4 replicates). (C) Electrophoretic mobility shift assays showing the nuclear extracts which contain Hif-1α2 could bind to the hypoxia response element (HRE) of DJ-1 promoter. Unlabeled probes were used as competitors. SH represents shift binding of Hif-1α2, UB represents unspecific banding. (D) Knockdown of DJ-1 promoted H2O2 accumulation, and reduced the ratios of reduced to oxidized glutathione (GSH/GSSG) in muscle during flight (n=5 replicates, 3 locusts/replicate), and (E) impaired long-term flight performance of locusts (n≥27). Student’s t-test for pairwise compariso