Prenatal hyperbaric normoxia treatment improves healthspan and regulates chitin metabolic genes in Drosophila melanogaster

Aging is a universal, irreversible process accompanied by physiological declines that culminate in death. Rapid progress in gerontology research has revealed that aging can be slowed through mild stress-induced hormesis. We previously reported that hyperbaric normoxia (HN, 2 atm absolute pressure with 10% O2) induces a cytoprotective response in vitro by regulating fibronectin. In the present study, we investigated the hormetic effects of prenatal HN exposure on Drosophila healthspan related to molecular defense mechanisms. HN exposure had no disruptive effect on developmental rate or adult body weight. However, lifespan was clearly enhanced, as was resistance to oxidative and heat stress. In addition, levels of reactive oxygen species were significantly decreased and motor performance was increased. HN stress has been shown to trigger molecular changes in the heat shock response and ROS scavenging system, including hsp70, catalase, glutathione synthase, and MnSOD. Furthermore, to determine the hormetic mechanism underlying these phenotypic and molecular changes, we performed a genome-wide profiling in HN-exposed and control flies. Genes encoding chitin metabolism were highly up-regulated, which could possibly serve to scavenge free radicals. These results identify prenatal HN exposure as a potential hormetic factor that may improve longevity and healthspan by enhancing defense mechanisms in Drosophila.


INTRODUCTION
All living systems have the intrinsic ability to respond to external and internal stressors and to maintain homeostasis [1,2]. A molecular homeodynamic balance increases the efficiency of multiple pathways, including nuclear and mitochondrial DNA repair, detoxification and immune and stress responses [3]. A failure in the homeodynamic pathway culminates in death because of aging due to accumulated molecular and cellular damage [2,4,5].
The phenomenon that low doses of potentially toxic substances can induce a beneficial biological response is termed hormesis [6][7][8][9]. These beneficial low doses, called mild stress, include a wide variety of biological, chemical, and mechanical stimuli. Accumulating studies have established a strong link between hormesis and aging, as a result of overcompensation of defense mecha-nisms induced by mild stress. For example, exposure to mild stresses such as irradiation, heat, or hypergravity improves mean lifespan in Drosophila melanogaster [10][11][12][13] and improved stress tolerance in D. melanogaster [12,14] and in Caenorhabditis elegans [15].
We previously reported that hyperbaric normoxia (HN, 2 atmospheres absolute with 10% O 2 ) as a mechanical mild stressor shows hormetic effects, such as inducing a cytoprotective response in vitro by regulating fibronectin expression and DNA damage [16]. Mechanical stress caused by stretching, compression, fluid shear stress, or hydrostatic pressure is the primary regulator to induce homeostatic mechanisms. It has been shown to influence major developmental processes, such as in skeletal muscle, bone, cartilage, blood vessels and heart, and to affect diverse cellular processes including cellular growth, differentiation, migration, apoptosis, AGING (Albany NY) and senescence [17,18]. Mechanical stress has been well studied in development research [19], but its application to biogerontologic research as a hormetic factor remains unclear.
There is growing evidence to suggest that exposure to environmental stress during critical periods of development result in permanent effects on behavior and metabolism, as well as growth, reproduction, stress tolerance and lifespan [20][21][22]. These observations were explained by the concept of developmental plasticity [23], which causes genetic or epigenetic modulation. In particular, the hormetic effects of dietary regulation on developmental plasticity have been investigated in lifelong health studies across many different animal species [24][25][26][27]. The effects of hormesis and its application to biogerontologic research are being explored, and one of the next challenges is to identify new mild stresses that have hormetic effects at the organismal level.
Hyperbaric normoxia stress shows no genotoxicity on various cell lines, and in fact shows cytoprotective effects [16]. However, evidence that these hormetic effects of HN also occur in vivo is lacking. In the present study, using D. melanogaster as a model system we investigated the effect of prenatal HN exposure on healthspan, including developmental rate, fecundity, severe stress tolerance, longevity and behavioral aging. In addition, we addressed the molecular consequences of HN, as a mechanical mild stress in response to the relevant aging trait.

Characterization of HN in D. melanogaster development
To determine the effects of HN in D. melanogaster, we examined whether prenatal HN exposure affected development at each stage of growth. We observed no significant effects of HN on growth compared to controls. Although the timing from larva to pupa was delayed at day 5 (p =0.003), the total larval growth pattern, total development time from egg to eclosion (Fig.  1A) and total average eclosion rate (egg to adult survival) did not differ between HN and control groups (Fig. 1B). Moreover, HN did not affect the mean body weight of males, females or larvae at eclosion (Fig. 1C) or wing size (Fig. 1D).

Protection against heat and oxidative stresses by prenatal HN exposure
We previously showed that hyperbaric normoxia protects normal fibroblasts (WI-38) against oxidative stress [16]. We therefore examined whether HN treat-ment also enhanced stress tolerance in D. melanogaster (Table 1A). To determine the tolerance to oxidative stress, flies were fed 18 mM paraquat added to standard food and their survival was monitored. We found that for both male and females, HN-treated flies exhibited a significant resistance to oxidative stress, 12% increase

AGING (Albany NY)
of median survival in males (Fig. 2C) and 47% in females (Fig. 2F), whereas control flies exhibited high sensitivity to paraquat. Under 40°C heat stress, both male and female HN-treated flies showed increased thermal tolerance by 20% in both males ( Fig. 2A) and females (Fig. 2D). In addition, when subjected to nutrient deprivation, only female flies showed increased resistance to starvation, compared to their controls (males, Fig. 2B; females, Fig. 2E). These results suggest that HN plays a protective role against different stresses in Drosophila, which is in agreement with previous in vitro reports [16].

Enhanced longevity and healthspan by hyperbaric normoxia
To gain further insight into the protective effects of HN, we assayed the lifespan in treated and untreated flies (Table 1B). We found that the median lifespan of HNtreated flies was increased by 12% in males (Fig. 3A) and by 14% in females (Fig. 3B). Motor performance is regarded as a marker of healthspan in D. melanogaster. [28,29]. To measure this, we tested the vertical climbing ability of treated and untreated flies. We found that HN treatment improved the flies' ability to reach

AGING (Albany NY)
the top of the vial (males, Fig. 3C; females, Fig. 3D), indicating an improved motor performance. We also tested reproductive output, which is often negatively correlated with extended lifespan and healthspan [30,31]. However, our results showed that while HN-treated female flies showed decreased fertility at day 8, there was no detrimental effect on average daily fecundity compared to controls (Fig. 3E). This implies that prenatal HN exposure has the ability to delay the physiological age without any associated reproductive costs.

Changes in ROS production and oxidative stressrelated gene expression
Aging leads to increasing ROS as well as impaired motor performance associated decreased lifespan [32]. The relationship between ROS and healthspan has been studied extensively. In the present study, we examined ROS levels in HN-treated flies, using a FOX assay. Compared to control flies, the average amount of ROS in HN-treated flies was decreased by 10% for both males and females (Fig. 4A). Otherwise, there was no difference in expression levels of mitochondrialencoded genes, such as cytochrome c oxidase subunits I and III (COXI, COXIII) and cytochrome b (Cytb) in males or females (Fig. 4B). To test if ROS reduction by HN exposure leads to altered gene expression associated with aging and oxidative stress, we monitored mRNA levels of candidates. We analyzed the gene expression level of Drosophila methushelah (mth) and Sir2, 4E-BP, FOXO and Tor in nutrient signaling, there was no changes (males, Fig. 5A; females, Fig.  5B). However, Oxidative stress-related genes including hsp70, catalase, glutathione synthase (GS), Jafrac 1, and Superoxide dismutase 2 (MnSOD) were upregulated following HN exposure (males, Fig. 5C; females, Fig. 5D). These data suggest that the prolonged healthspan that results from prenatal HN exposure is associated with anti-oxidant defense mechanisms.

Global gene expression profiles in HN treated healthspan increased flies
To obtain a global view of the genomic response to HN treatment in the healthy aging phenotype in male and female flies, we performed a microarray using 32,162 probe sets and analyzed it together with its control. Genes with a more than 1.5-fold change (with statistical significance) following HN treatment were presented as Venn diagrams (Fig. 6A). In total, 4,293 genes (3,946 induced; 347 repressed) were altered in males and 1,841 genes (1,163 induced; 678 repressed) were altered in females following prenatal HN exposure. Comparison of these two sets revealed a significant overlap of 473 genes, among which 441 genes were up-regulated in the HN-treated group. These up-regulated genes can be analyzed in Gene Ontology (GO) which is grouped into three major categories, including biological process (BP), cellular component (CC), and molecular function (MF) (Fig. 6B). The top five sub-categories in BP were indicated in aminoglycan metabolic process (18 genes), chitin metabolic process (17 genes), Polysaccharide metabolic process (7 genes), myofibril assembly (7 genes) and ciliary or flagellar motility (5 genes).

Chitin metabolism is involved in the improved healthspan seen in HN-treated flies
Our microarray analysis revealed changes in 18 transcripts involved in aminoglycan metabolic process. Among these, 17 transcripts were found to be involved in chitin metabolism, which is related to aminoglycan metabolism according to functional enrichment map. Using KEGG pathway analysis to confirm the over-lapping upregulation following prenatal HN exposure, amino sugars and nucleotide sugar metabolism were enriched in the HN-treated group. Potential healthspan-promoting genes Chitinase 6 (Cht6: CG43373), Chitinase 7 (Cht7: CG1869) and kkv, were selected for further analysis. Validation by qPCR revealed that mRNA levels of Cht6, Cht7, and krotzkopf verkehrt (kkv) were highly increased in prenatal HN-exposed flies (males, Fig. 7A; females,  Fig. 7B). These observations identify chitin metabolism as a potential molecular mechanism involved the hormetic effects of HN that result in a prolonged healthspan.

DISCUSSION
The aims of this study were to test the potential of HN as a mild stressor that can prolong lifespan and healthspan in Drosophila and to elucidate the hormetic mechanisms by which HN promotes longevity. While the hormetic effect of exposure to various mild stresses on healthspan and longevity is well documented at adult stages, the potential hormetic effects of prenatal exposure to mild stress remain relatively unknown. Most of the prenatal stress studies were restricted to harmful environmental exposures such as carbon monoxide, alcohol and drugs, which can lead to subvert normal development [33,34]. In general, prenatal exposure to mild stress has a more pronounced impact on development and the whole life of an organism than postnatal exposure. Therefore, in the present study, before using prenatal HN exposure as a mild stressor, it was necessary to determine any side-effects on the development of D. melanogaster. Our data showed that prenatal HN exposure had no detrimental effects on development, such as larval growth rate and embryonic mortality, and body weight of larvae and adults. In addition, prenatal HN exposure did not affect wing size, which is used as an indicator of body size, after growth has stopped. Therefore, our observations indicate that prenatal HN treatment has no deleterious effects on Drosophila development.
Mild stress intervention has been shown to improve stress resistance and lifespan, and therefore slow down the aging process [35,36]. Long-lived mutants often show increased stress resistance to environmental stimuli [35]. Therefore, we hypothesized that HNinduced mild stress would enhance lifespan and stress resistance against critical heat, starvation and oxidative stress. Our results showed that prenatal HN exposure induced longevity and stress resistance to thermal and oxidative stress in both male and female flies. We found no change in starvation resistance, but this could be correlated with a number of other life-history traits, in-cluding body size. Thus, our study provides additional independent evidence for the use of HN-induced mild stress as a hormetic effector in Drosophila, as was suggested by a previous in vitro study [16].

AGING (Albany NY)
Simply extending lifespan has no value if it is not coupled with improved healthspan. Therefore, sought to determine whether the beneficial effects of prenatal HN exposure could be explained in terms of improved and prolonged lifespan. In Drosophila, muscle strength is essential for motor performance and is known to decline with age, causing age-related health problems [28,29]. Locomotor activity is therefore a reliable marker of healthspan. Indeed, we found that HN exposure improved locomotor activity, indicating an improvement in the healthspan of the treated flies. We also tested whether prenatal HN exposure affected female egg-laying ability. Previous diet manipulation studies have reported that an increased lifespan is accompanied by compromised reproduction [30,31]. The existence of such a trade-off between longevity and fecundity seems reasonable in terms of allocation of resources under food-limited conditions. However, reduced reproduction does not necessarily improve lifespan, and several studies have reported no correlation between longevity and fecundity [37][38][39]. In the present study, we showed that prenatal HN exposure increased stress tolerance and healthspan without a concomitant reduction in fecundity.
It has been proposed that prenatal HN-induced stress triggers a defense mechanism related to improving longevity and healthspan. The free radical theory of aging suggests that ROS-induced accumulation of damage to cellular macromolecules is a major determinant of lifespan [40]. To this end, we assessed the levels of cytosolic reactive oxygen species (ROS) and found a decrease in cytosolic ROS production following HN treatment. It will be interesting to determine how prenatal HN stress induces the defense mechanism against ROS. One type of defense mechanism involved in hormesis is the heat shock response. Hsp70 is a stress-induced chaperone that regulates lifespan in flies [14,41], and its expression is reduced with increasing age in D. melanogaster [42]. Another type of intervention used to increase healthspan and decrease oxidative stress is the ROS scavenging system. Superoxide dismutase and catalase provide the main enzymatic antioxidative defense in insects. During aging, expression levels of MnSOD are down-regulated in Drosophila [43,44]. MnSOD is localized to the inner mitochondrial space and plays a role in mitochondrial DNA protection by converting superoxide to hydrogen peroxide and oxygen. MnSOD knockdown has been shown to reduce oxidative stress resistance and lifespan in flies [45], as well as in a muscle tissue-specific knockdown [46,47]. Catalase converts hydrogen peroxide to water and oxygen and mitochondria-specific catalase overexpression was shown to extend the lifespan of mice [48]. In Drosophila aging gut, the catalase expression level is improved by rapamycin [49] and catalase activity is diminished during aging in C. elegans [50].
Although we obtained some molecular evidence to link antioxidant defense and aging, the underlying mechanism of the healthy aging phenomenon following HN stress is still not clear. We therefore carried out a genome wide microarray analysis, and observed enrichment for genes related to chitin metabolism in the prenatal HN-treated flies. Changes in chitin metabolism-related genes would be predicted to result in an accumulation of chitin. Chitin oligosaccharides are a major component of the inset cuticle and are known to act as potential antioxidants by scavenging ROS [51]. Chitin is synthesized by chitin synthase enzyme coding gene kkv and changes in its activity have been related to aging in D. melanogaster [52,53]. Interestingly, the interaction partners for chitin synthases, or chitinases, are unknown in insects [54]. We propose that HN stress acts as a mechanical cue to stimulate chitin metabolism as part of an antioxidative defense mechanism.
In summary, we have identified HN stress as a highly effective hormetic factor. We provide independent evidence for the role of prenatal exposure to HN in lifespan extension and increased stress tolerance without detrimental effects on development in both male and female flies. We also found that prenatal HN exposure enhanced motor performance, indicating an improvement in healthspan as well as lifespan. Interestingly, reproductive ability was not affected by HN stress, which is generally considered a resource trade-off for healthspan extension. Furthermore, HN stress decreased ROS levels by inducing the antioxidant defense mechanism. Considerable overlap between phenotypical and molecular mechanisms was explained using a genome-wide analysis of gene expression profiles. Strong up-regulation of genes involved in chitin metabolism was detected, which could possibly function in the process of scavenging free radicals. These results suggest that prenatal HN exposure is a major hormetic factor that improves longevity and healthspan in Drosophila by enhancing antioxidant defense mechanisms. Further analyses will be required to understand how the hormetic effects of HN promote healthy aging in other living systems.

Fly stocks and maintenance
Wild-type Drosophila melanogaster w 1118 were obtained from the Bloomington Drosophila Stock Center (Bloomington, IN, USA) and maintained in a climate-controlled incubator with 25°C and 50 ± 5% relative humidity on 12:12 h light/dark cycle. All flies AGING (Albany NY) used in this study were raised on fresh medium containing 6% cornmeal, 3% yeast, 3% sugar, 6% glucose, and 0.6% agar.

Experimental design for hyperbaric normoxia incubation
All hyperbaric normoxia (HN; 2ATA in normoxic conditions) exposures were carried out in a custommade animal incubator designed for pressure control [55] without change of temperature (25°C), humidity (50 ± 5%) and under a 12:12 h light/dark cycle. Fertilized eggs from 30 3-day-old pairs were collected in batches of 30 into vials, then exposed to HN until eclosion. After HA exposure all newly born (F1) flies were maintained in a climate-controlled incubator as described above.

Measuring developmental time, eclosion rate and fecundity
We measured larval development time from egg to pupa (at 24-h intervals), eclosion rate from egg to adult and weight of larvae and adults (male or female). For fecundity, newly eclosed F1 generations of 15 virgin female and 7 male flies were collected and placed into a vial. Flies were transferred to fresh vials every day and the number of eggs laid in a 24-h period was counted for 15 days with ten replicates.

Body weight and wing size analysis
Body weight of individual larva and adult flies was measured. Flies were reared under the same growth conditions and were age-matched (2 days old) before weighing.
To estimate body size, variation in wing size was measured. The left wing was removed and mounted onto double-sided tape on glass slides and sealed with a cover glass. Images of the wings were taken using a Leica MZ 125 microscope and Leica Application Suite LAS version 3.7.0 software (both from Leica Microsystems).

Climbing and longevity assays
The climbing assay was performed as previously described by Ganetzky et al. [28] with some modifications. Ten-day-old flies (20 males or 20 females per vial) were collected and placed into a vial (height, 10 cm). After a 1 h recovery period from anesthesia, each vial was gently taped to the bottom. At this point, we counted the number of flies that climbed up to a vertical distance of >7.5 cm within 5 s or 7 s. This procedure was repeated five times. After five trials, the number of flies in each vial was counted. At least 100 flies were used for each condition tested.
Lifespan assay was performed as described previously [56]. In short, 80 male or female flies (1-3 day after eclosion) were cultured in an aging chamber [57]. The number of dead flies was counted every 2-3 days, when fresh food vials were replaced. This procedure was repeated three times.

Assays for stress resistance of adults
To examine ROS tolerance, flies were placed in a vial containing fresh food and 18 mM paraquat (methyl viologen; Sigma-Aldrich, USA), after wet-starvation in a vial containing 1% agar for 6 h. The number of dead flies was recorded every 3 h until all flies had died. Similarly, starvation resistance was measured by feeding with only water in a vial containing 1% agar (Sigma-Aldrich, USA). Dead flies were counted every 3 h. To test heat tolerance, flies were kept in a 40°C incubator and paralyzed flies were counted every 10 min. Each experiment was performed with 10-day-old flies (20 males or 20 females per vial) with at least three replicates.

Quantitative real-time PCR analysis
F1 generation flies were frozen in liquid nitrogen and stored at -80°C until analysis. Total RNA was extracted from homogenized whole-body lysates using RNAiso reagent (TAKARA, Japan). After treating with DNase I (Invitrogen, USA), cDNA synthesis was performed using a PrimeScript RT Reagent Kit (TAKARA, Japan). Quantitative real-time PCR was performed using Power SYBR Green (Applied Biosystems, USA) and StepOnePlus Real-Time PCR System (Applied Biosystem, USA). To determine mRNA expression levels for each gene, C T values were normalized to those of rp49. Mean fold induction was calculated from the values from 3-6 independent experiments.

Quantification of mtDNA
Total DNA was extracted from 10 day old flies from the F1 generation using a G-spin Genomic DNA extraction kit (iNtRON Biotechnology, Korea) and subjected to quantitative real-time PCR as described above. Copy number of mtDNA was estimated using Cox I, Cox III, and Cyt b primers [58], relative to rp49.

ROS measurement
The rate of ROS production in 20 ten day old flies was measured using a ferric-xylenol orange peroxide assay [59]. Whole bodies were minced in 250 µl distilled AGING (Albany NY) water containing aminotriazol (2 mg/ml). Experiments were performed in triplicate. After centrifugation 40 µl of supernatant was mixed with 360 µl FOX reagent (250 µM Fe(NH 4 )2SO 4 ·6H 2 O, 25 mM H 2 SO 4 , 4 mM butylated hydroxyl toluene in HPLC grade methanol, 100 µM xylenol orange) containing 100 mM sorbitol (all from Sigma-Aldrich, USA). After incubation at room temperature for 40 min, absorbance of the ferricxylenol orange complex was estimated at 560 nm using a SpectraMax microplate reader (Molecular Devices, USA).

Microarray
Total RNA from 20 male and 20 female flies from control and HN-treated groups were collected as described above. For each RNA, the synthesis of target cRNA probes and hybridization were performed using a LowInput QuickAmp Labeling Kit (Agilent Technologies, USA) according to the manufacturer's instructions. Briefly, 25 ng total RNA was added to the T7 promoter primer mix and incubated at 65°C for 10 min. cDNA master mix (5X First strand buffer, 0.1 M DTT, 10 mM dNTP mix, RNase-Out, and MMLV-RT) was added to the reaction mix. Samples were incubated at 40°C for 2 h, and RT and dsDNA synthesis was terminated by incubating at 70°C for 10 min. Transcription master mix was prepared according to the manufacturer's protocol (4X Transcription buffer, 0.1 M DTT, NTP mix, 50% PEG, RNase-Out, inorganic pyrophosphatase, T7 RNA polymerase, and Cyanine 3-CTP). Transcription of dsDNA was performed by adding the transcription master mix to the dsDNA reaction samples and incubating at 40°C for 2 h. Amplified and labeled cRNA was purified on an RNase mini column (Qiagen, USA) according to the manufacturer's protocol. Labeled cRNA target was quantified using an ND-1000 spectrophotometer (NanoDrop Technologies, USA). After checking labeling efficiency, 1650 ng of each cyanine 3-labeled cRNA target was used for cRNA fragmentation by adding 10X blocking agent and 25X fragmentation buffer and incubating at 60°C for 30 min. Fragmented cRNA was resuspended with 2X hybridization buffer and directly pipetted onto an assembled Agilent Drosophila (V2) Gene Expression 4X 44K Microarray. Arrays were hybridized for 17 h at 65°C in an Agilent Hybridization oven (Agilent Technologies, USA). Hybridized microarrays were washed according to the manufacturer's protocol (Agilent Technologies, USA).

Microarray data acquisition and GO analysis
Hybridization images were analyzed using an Agilent DNA Microarray Scanner (Agilent Technologies, USA) and data quantification was performed using Agilent Feature Extraction software 10.7 (Agilent Technologies, USA). Average fluorescence intensity was calculated for each spot and local background was subtracted. All data normalization and selection of fold-changed genes were performed using GeneSpringGX 7.3.1 (Agilent Technologies, USA). Normalization for the Agilent one-color method was performed: Data transformation, set measurements less than 5.0 to 5.0; Per Chip, normalize to 50th percentage. Averages of normalized ratios were calculated by dividing the average of control normalized signal intensity by the average of test normalized signal intensity.
Functional annotation of genes was performed according to DAVID annotation tools [60,61]. Default DAVID parameters were used. Venn diagrams from two gene list (male and female) in microarray data analysis were created by R software package. All raw data were deposited in the NCBI Gene Expression Omnibus under accession number GSE84069.

Statistical analyses
Statistical analyses were conducted using PRISM version 6 (GraphPad Software, Inc., USA). All data except survival curves were presented as mean ± SEM and were statistically evaluated using Student t-test distribution. Results of lifespan and stress resistance studies were analyzed using the Kaplan-Meier survival test. Comparisons among groups were conducted using Gehan's Wilcoxon test.