During a space mission, fish embryos are exposed to two types of altered gravity. The first is hypergravity (∆g > 1 g) during launcher acceleration. During this 10-min take-off phase, gravity could be 3 g<∆g < 5 g depending on the thruster power of the launcher. The degree of acceleration varies according to the velocity of the launcher, the fuel burned by the engine, and the mass of the rocket. A study on the larvae of cichlid fish (Oreochromis mossambicus) and juveniles of the swordtail (Xiphophorus hellerii) looked at fish behavior, neuronal tissue and the otolith in these conditions, finding that fish have hypergravity perception (Brungs et al., 2011). In a permanent 3 g environment, developed fish showed a slight lateral tilt during swimming, probably due to asymmetric otolith growth (Hilbig and Anken, 2017). However, the influence of hypergravity on fish embryo development is poorly documented. In research investigating otolith formation, fertilized zebrafish (Danio rerio) eggs were reared at 3 g for 7 days. In the control group, by 16 h post fertilization (hpf) at 1g, precursors of the otoliths were observed. Zebrafish reared at 3 g from 1 to 7 days after fertilization exhibited significantly slower otolith growth than the controls at 1 g (Wiederhold et al., 2003). At 24 hpf, decreased pigmentation was observed when fertilized zebrafish eggs were under 3 g gravity. One day later, the difference with the control batch (1 g) disappeared (Aceto et al., 2015). A study with zebrafish egg and post-larvae under 7 days at 3 g exposure showed changes in their swimming and flotation patterns. Hatching rate showed a significant delay between 2- and 3-day post fertilization. Moreover, long-term hypergravity impacts the gene expression in genes implicated in epigenetic mechanisms (Salazar Moscoso et al., 2022). Medical research has applied long-term hypergravity (several days) to aquarium fish models to better understand vertebrate physiology. However, in a real space mission, this environment lasts only 10 minutes—the time to reach LEO. Therefore, this shorter duration of hypergravity was used for our study.

When ∆g < 1g, organisms are under microgravity. In a space mission, gravity is 10⁻³ g<∆g < 10⁻⁵ g (Black et al., 1995), whereas on the Moon it is ∆g = 0.166 g. In medical research, larval or juvenile fish have been used as animal models in LEO to investigate the development of the vestibular system and otolith in order to gain a better understanding of the human inner ear (Anken et al., 2016). In this context, ground-based experiments have been conducted to simulate microgravity using a drop tower (Anken and Hilbig, 2009), parabolic flight (Anken et al., 1998; Endo et al., 2002) or a clinostat/rotating wall vessel (Brungs et al., 2011; Edsall and Franz-Odendaal, 2014).

Multicellular life on Earth appeared around two billion years ago. The gravity on Earth is 1g, influencing the phylogenic development of living organisms. Compared to terrestrial vertebrates, aquatic organisms are subject to gravitational force and an opposing buoyancy force. This results in an aquatic organism feeling a reduced body weight compared to a terrestrial organism. This is the reason astronauts train in pools or simulate extravehicular activities in marine environments (Trout and Bruchey, 1969). The aquatic environment suggests that fish embryo development may be less impacted by microgravity than hypergravity in altered gravity conditions.

Determining the feasibility of space aquaculture requires switching from an aquarium fish model to an aquaculture fish model to conduct experiments on their biology in space (Przybyla, 2021). A journey to the Moon is the first step to verify before planning a potential aquaculture system for a future lunar base. This requires studying embryo/larvae survival rates, behavior and biological status in the conditions experienced during a lunar trip in order to optimize fish welfare.

Cortisol, a steroid hormone belonging to the glucocorticoid family, is involved in the regulation of various physiological mechanisms, including inflammation, metabolism and stress response. In fish, as in many other vertebrates, cortisol is considered the main stress hormone (Sadoul and Geffroy, 2019). It binds to glucocorticoid receptors (GRs) and mineralocorticoid receptors to activate or repress a number of genes, contributing to physiological adaptation following stress. The European seabass possesses two glucocorticoid receptors: GR1 and GR2. Expression of genes encoding these proteins has been detected early in an individual’s development in the species and used as a stress marker in various studies (Di Bella et al., 2008; Pavlidis et al., 2011; Tsalafouta et al., 2014; Geffroy et al., 2021).

Heat shock proteins are involved in animal development (Rutherford and Lindquist, 1998), as well as in fish stress response (Palmisano et al., 2000). In European seabass, Hsp90α is a chaperone protein for both GR1 and GR2 and plays a key role in regulating the availability of GRs for cortisol. Hsp90α activity has mostly been described during stress experiments related to exposure to warm temperatures, including in the seabass (Goikoetxea et al., 2021) and the gilthead bream Sparus aurata (Madeira et al., 2016). Monitoring GR1, GR2 and Hsp90α during altered gravity experiments can give us useful information on the development and the stress level of larvae hatched in microgravity or after being subjected to a hypergravity environment during embryo development.

Fish embryos and larvae have an efficient defense system to deal with aquatic pathogens, although they lack an adaptive immune system (with memory, and thus specific to the individual), which becomes functional at a later developmental stage. To recognize and efficiently destroy pathogens at an early stage of development, the innate immune system (without memory and non-specific) is activated via different pathways (Magnadóttir, 2006; Uribe et al., 2011). The key component of the innate immune system is the complement component 3 protein (C3), which plays a central role in the lytic and pro-inflammatory pathways (Yano et al., 1988; Najafpour et al., 2020). This protein was identified in the digestive system of Atlantic halibut larvae (Hippoglossus L.) as early as the onset of first feeding, 50 days after hatching (Hermannsdottir et al., 2009). Complement proteins have also been shown to be involved in tissue and organ regeneration. Distinct complement-triggered pathways have been shown to modulate critical responses that promote tissue reprogramming, pattern formation and regeneration across phylogenesis (Mastellos et al., 2013). Thus, monitoring C3 protein could provide information on the development of the non-specific immune system in altered conditions, assuming that environmental changes may have an impact on this.

Interleukin-1β (IL-1β) expression has been reported in several studies on vertebrates during spaceflight (Crucian et al., 2014; Buchheim et al., 2019; Krieger et al., 2021). It is an initiator of inflammation, and its expression in vertebrates can also be a response to stress (Brydon et al., 2005; Serradell et al., 2020). IL-1β has been studied in 13 teleost fish species, and it seems to function in a similar way as in other vertebrates including mammals (Uribe et al., 2011). Given that inflammation or stress can cause tissue damage affecting various organs and physiological systems of fish, in our study we quantified IL-1β transcripts to determine if simulated altered gravity impacts the expression of this innate immune gene in D. labrax embryo development.

Our study focused on the hatching rate and oxygen consumption of aquaculture fish eggs during a ground-based experiment simulating hypergravity using a centrifuge and simulating microgravity using a random positioning machine (RPM). The parameters studied after exposure to altered gravity were the RNA integrity number (RIN), the relative expression of glucocorticoid receptors (GR1, GR2), and Hsp90a in order to assess the stress level after conditions based on a real space mission. The C3 and IL-1β transcripts aimed to assess the early immune system status after exposure to both altered conditions just after hatching. The aim of this study was to lay the groundwork for a proof of concept of the feasibility of hatching aquaculture fish in LEO or on a future lunar base and to anticipate any countermeasures that may need to be implemented for fish welfare during such a mission.

To conduct the experiment, a polyvinyl chloride experiment tank was built to sit inside a CubeSat (a miniaturized satellite), which is 10 cm × 10 cm x 10 cm (Figure 2A/Figure 2B). The tank diameter was 6.4 cm and it weighed 639 g when filled with water; the tank volume was 230 mL. The altered gravity experimental groups were then hermetically sealed into the CubeSats and the control groups were sealed in bottles without the possibility of renewing the water or injecting oxygen. While the experimental CubeSat can carry 200 European seabass eggs (the industrial protocol is 1 egg/mL), for this experiment on altered gravity, we decided to reduce the number of samples to 100 eggs to ensure a comfortable environment (0.43 eggs/mL), especially in terms of oxygen availability for the entire week of testing.

A week before the experiment, four male and four female Dicentrarchus labrax were selected at the Ifremer aquaculture station (Palavas-les-Flots, France) based on sperm availability and ovulation stage. The selected fish were then isolated in a recirculating aquaculture system (RAS) regulated at 13.2 °C. Two days before the date of the scheduled fertilization, the females were injected with gonadotropin-releasing hormone (10 μg/mL). The day of fertilization, the females were ready for egg collection. Only one male and one female with the best oocyte quality were selected for controlled fertilization in order to produce a single family for the hypergravity experiment (HGE) and the simulated microgravity experiment (SMGE) to minimize parental influence on the results. A volume of 2 mL of oocytes (n ≈ 2000) was sampled and placed into a small 150-mL beaker, assessed under the microscope and sprayed with 1 mL of fresh sperm. The fertilized eggs were then placed into an aquatic incubator at 14.5 °C, pH ≈ 8.15 and salinity ≈38%. Viable eggs were collected to be sampled.

The rate of oxygen consumption of fish embryos is difficult to measure, due to the small size of eggs and low rates of oxygen uptake. The oxygen concentration inside the CubeSat modules was regularly measured using non-invasive optical fiber technology. This method avoided opening the sealed CubeSat with the risk of degassing, preserving the fish embryos from possible stress. By placing the edge of the fiber in front of the sensor (Figure 2E), the transmitter is able to analyze the received signal and translate it into an oxygen rate (in % of air saturation) readable directly by Presens FiBox software (Figure 2F). Calibration was required before the measurement, using Milli-Q water at 0% and 100% of oxygen saturation. The temperature of the module was recorded for each measurement because this influences the calculation of oxygen concentration. All modules were filled at 22 hpf with the same marine water with constant salinity (34.8 PSU). A measurement with filtrated water (0.2 µm to avoid bacteria presence and respiration) and without eggs showed that the CubeSat lost about 15% of oxygen, probably due to a porous gasket. This loss was considered in the calculation of oxygen consumption. Oxygen measurement was performed for all groups just after the CubeSat module was hermetically sealed at 22 hpf.

Embryo oxygen consumption was calculated for the SMGE and control following the formula below:

A: [Biological oxygen loss] (µgO₂/Egg) ∆ O 2 22  hpf − 70  hpf − ∆ O 2  ∗  15 % mechanical loss  ∗  0.23 CubeSat volume (1)

B: [Number of hours since CubeSat sealing] ∆ 22  hpf − 70  hpf (2)

Oxygen consumption during the microgravity experiment (µgO₂/eggs/h) was calculated with A/B.

Total RNA was extracted from the larvae subjected to hypergravity conditions, simulated microgravity conditions and control conditions using the Nucleospin ^® RNA Plus XS kit (Macherey-Nagel, Düren, Germany). RNA integrity analysis was performed using RNA ScreenTapes and a TapeStation 4,150 device (Agilent, Santa Clara, California, United States). From each sample, 200 ng of total RNA was reverse transcribed using random primers, dNTP, RNaseOut and MML-V reverse transcriptase (all from Invitrogen, Cergy Pontoise, France) following the manufacturer’s instructions.

As the hatching rate data was expressed in percentages and close to the extreme [0; 1], statistical analysis was performed with a converted value using y = a r c ⁡ sin x . The transformed data was subjected to one-way statistical analysis of variance (ANOVA). GraphPad Prism 9.0 software (Dotmatics, Boston, Massachusetts, United States) was used to perform statistical analyses on biomarkers. Homogeneity of variance was determined using Fisher’s test and the normality of distribution was determined using the Shapiro–Wilk test. When homogenous variance and distribution were observed, the parametric Student’s t-test was used. When variance and distribution were not homogeneous, a logarithmic transformation was applied to the data to make it statistically in accordance with the test requirement. For the oxygen measurement, the slopes of the trend curves were compared using a linear regression method. Minimum level of significance was p < 0.05 and triplicate results were shown as mean value ±SD.

From egg fertilization to the end of the experiments, the water incubation temperature was kept at 14.5^±1.2 °C in the hypergravity (HGE) conditions and 16.5^±1.8 °C in the microgravity (SMGE) conditions. This temperature difference was due to the absence of thermoregulation in the SMGE room, and it is likely to explain why larvae were observed and counted at different times for HGE (at 96 hpf) and SMGE (at 72 hpf). Embryogenesis time is directly related to incubation temperature. Devauchelle and Coves (1988) put forward a robust model to predict European seabass egg hatching date based on incubation temperature, and this model has been regularly confirmed (Saka et al., 2001; Cucchi et al., 2012; Przybyla et al., 2020). The model accurately predicted the HGE hatching time within 1 hour. In the SMGE conditions, the European seabass eggs hatched 6 h before the expected time, which is in line with previous observations of killifish in microgravity (Scheld et al., 1976; Lychakov, 2016).

Hatching rates in other species have been previously evaluated during LEO missions: for example, with aquarium fish (n = 50) as part of the Skylab3 mission the hatching rate was 96%, and with killifish (n = 50) on the Apollo-Soyuz mission it was 22% (Hoffman et al., 1977; von Baumgarten et al., 1975). In later missions, the hatching rates recorded for medaka were 88% (n = 48) and 25% (n = 4) (Ijiri, 1998; Niihori et al., 2004). Results from our 10-min 5 g hypergravity experiment with European seabass eggs showed no significant difference compared with a control reared under optimal industrial conditions (Figure 3). The group exposed to simulated microgravity for 39 h recorded the same range of hatching performance as the industrial control group. Because we had access to only one RPM, only two hatching rates were assessed (exposed and control) for the simulated microgravity experiment, so this result should be considered to indicate a tendency rather than as robust data. For an unexplained reason, the simulated microgravity control group in the RPM room had a bad hatching percentage, which cannot be related to parental genetic inheritance because all the embryos were siblings.

The initial water used to fill the CubeSat containing the fish eggs was over-oxygenated to ensure a sufficient oxygen level for the entire experiment (50 h) in total confinement in the CubeSat. Despite the precautions taken, the initial saturation was slightly different between the exposed group (248.7%) and the control group (240.2%) at the beginning of the experiment. Nonetheless, oxygen concentration inside the two sealed tanks was never under 146% of air saturation, and the cause of mortality of any unhatched eggs in the experiment cannot be attributed to lack of oxygen. Oxygen concentration kinetics during confinement before exposure and during the microgravity experiment are presented in Figure 4.

The slopes of the trend curves of the exposed group and the control group were not significantly different (p = 0.8074), suggesting that the microgravity environment did not generate an additional biological demand in oxygen during embryonic development. Furthermore, a calculation of individual oxygen consumption based on the oxygen volume consumed during the experiment and the surviving fish (live embryos and hatched larvae) in the two groups revealed that the embryos in microgravity did not consume more oxygen than the control group. The number of dead eggs (thus unhatched) was higher in the control group (52.08%) than in the exposed group (11.58%). We were unable to check when exactly the mortality occurred, before or during the microgravity exposure. Oxygen consumption values in the SMGE were consistent with those observed in a previous study with Dicentrarchus labrax (Pionetti, 1983) and other marine species (Table 5). Statistically, individual egg oxygen consumption did not differ between exposed and control groups in simulated microgravity.

In order to assess possible degradation during altered gravity exposure, the RNA integrity number (RIN) was estimated. RNA plays a central role in reading the information encoded in the genome and transferring it into proteins. We used the RNA integrity number (RIN) as a diagnostic tool to estimate the quality of the samples, as it has been successfully used as an indicator of good RNA transcription and larvae development at hatching in other studies (Schroeder et al., 2006). RIN values range from 10 (intact) to 1 (totally degraded). Figure 5 shows that 10 min spent at a hypergravity of 5 g slightly but significantly decreased the RNA integrity number (p-value = 0.009), while 39 h of simulated microgravity did not significantly affect it.

In this study, all RIN values obtained were above 7.0, which is considered to be RNA of high quality in the zebrafish Danio rerio (Mazzolini et al., 2018), red seabream Pagrus major (Nakatsuji et al., 2019) and gilthead seabream Sparus aurata (Vallejos-Vidal et al., 2022). While the hypergravity experiment showed a significant influence on RNA integrity, the RIN mean value (7.6 ^±0.4) stayed in the high-quality range and was not different from the control of the microgravity experiment. This observation correlates, from a physiological point of view, to the good hatching rate in the HGE group (exposed 81.97%^{± 5.96} and control group 90.13%^{± 4.36}), suggesting that the environmental change from a hypergravity of 1g–5 g in 10 min for embryos at 49 hpf had no deleterious impact on hatched larvae of European seabass. In the microgravity experiment, the RIN mean value was not significantly different from the control, suggesting that mRNA integrity and the transcriptomic mechanism in European seabass embryos were not altered by 39 h of simulated microgravity from 34 hpf to hatching time at 96 hpf. This observation correlates to the good hatching rate in the SMGE group (88.42%).