All the procedures were approved by CCMAR Animal Welfare Committee (ORBEA CCMAR-CBMR) and Direcção-Geral de Alimentação e Veterinária (DGAV) of the Portuguese Government, according to National (Decreto-Lei 113/2013) and EU legislation (Directive 2010/63/EU) on the protection of animals used for scientific purposes. In addition, protocols were approved by institutional Animal Care Committees at each of the Canadian Universities where authors of this paper are based. Procedures were only applied to live animals by authorized users.

Cuttlefish (S. officinalis) were reared according to the latest culture technology described in Sykes et al. (2014). All experiments used juveniles produced from eggs laid by a F6 captive stock. The mass of group 1 was 59.3 ± 5.4 g (N = 18) and of group 2, used only in mantle oxygen consumption experiments to avoid limiting oxygen diffusion across thick mantle tissue, was 1.90 ± 0.14 g (N = 7). Experiments were done during May 2016, at CCMAR's Ramalhete Aquaculture Station (Ria Formosa, Portugal—37°00′22.39″N; 7°58′02.69″W). Temperature, salinity, and DO₂ were measured daily, at 9 h 30, in the stock tank. Both temperature and DO₂ were measured with a VWR DO220 probe, while salinity was measured with a VWR EC300 salinity meter. Water temperature was 20.5 ± 1.27 (S.D.)°C, salinity was 35.9 ± 0.9 (S.D.) g L⁻¹ and DO₂ level was 101.1 ± 1.2 (S.D.)%. Cuttlefish were fed frozen grass shrimp (Palaemonetes varians) ad libitum on a daily basis.

Two closed respirometry (Lefevre et al., 2016) acrylic chambers (48.5 × 28.5 × 29 cm), each with an external recirculating pump (Maxi-Jet PH MP400) and a PASCO PasPort PPS-2169 multiparameter sensor were used. The recirculating system had a total volume of 41 L. Seawater entered and exited on opposite sides of the chamber at a flow of 200 L h⁻¹. Dissolved oxygen saturation data were collected in the upper right section of the chamber at 1 Hz in the PASCO Capstone software and at a temperature of 19.9 ± 1.36 (S.D.) °C. The chamber was completely isolated from any visual stimuli (both researchers and light) by the application of matte black stickers on the outer walls, except from above and one of the 48.5 cm sides. This side had a rectangular LED stripe (50 × 30 × 50 × 30 cm) attached that emitted red light (wavelength peak of 630 nm) until the chamber was completely filled with water and, thereafter, white light (wavelength peaks at 460, 520 and 630 nm) with an intensity of 80 lux. These conditions allowed for the recording of movement (from above) with a Logitech C615 HD camera, and of behavior and ventilation rates (from the side) with a GoPro Hero 3⁺ Black Edition. Video recordings were made for the duration of each experiment at 1,920 × 1,080 pixel resolution and 24 frames per s (fps), from above, and 1,920 × 720 pixel and 50 fps, from the side.

Initially, 6 animals were individually exposed to a DO₂ of 100% and 6 other animals to a DO₂ of 50% oxygen saturation for 60 min. Each condition was tested pairwise, 1 cuttlefish at 100% DO₂ vs. another at 50% DO₂. At the end of this procedure all the animals were euthanized. Samples of mantle, gill, and heart were rapidly removed and frozen in liquid nitrogen for biochemical analysis. Rates of oxygen consumption ṀO₂ were determined for animals exposed to 50% DO₂ during 60 min. In addition, 6 further control animals (DO₂ of 100%) were used to collect ṀO₂ during 90 min, to increase the number of available data points for regression fitting. Background ṀO₂ (resulting from consumption of bacteria and electrode) was also assessed with no animal (N = 3) in the chamber for 90 min and was found to be negligible.

Hypoxia at 50% DO₂ was achieved by gassing seawater with pure N₂ using a ceramic stone in a closed 90 L reservoir. This water was then injected into the chamber, until the 41 L respirometry recirculating system was filled, and where the animal was already resting. Animals were removed from the stock tank with a black net and transported in a 10 L black bucket filled with seawater to the weighing station. They were individually weighed in a black plastic tray and quickly transported to the chamber area of the experimental room which was illuminated with 1 lux red light conditions (Philips TL-D Colored 18W Red 1SL/25). An outer room containing the data collection stations was isolated from the holding chambers. To avoid inking in the experimental chambers, animals were left resting in the bucket for 5 min before being lightly sedated (mild anesthesia) by adding ethanol to a concentration of 2.5% (Fiorito et al., 2015). After sedation (≈2 min), animals were transferred to the chamber, which had 2 cm in height of seawater and thereafter filled with seawater from the 90 L reservoir (either 100 or 50% DO₂). As soon as the chamber was filled (≈2 min) and air purged, the system was put in recirculating mode, light switched to white and data acquisition was started on all the probes. This filling time was enough for recovery from sedation at this concentration (Gonçalves et al., 2012).

Ventilation rates were determined from video recordings, according to the following sampling scheme for each animal: (a) first and last 10 min—10 samples of 1 min each; b) from 10 to 50 min in hypoxia and normoxia (control)—1 sample every 10 min. After the procedure, the chamber was emptied, the animal removed, and euthanized with 10% ethanol in seawater (Sykes et al., 2012), the brain was then bisected downwards and forwards, followed by two lateral cuts to sever the brain from the optical lobes (Lewbart and Mosley, 2012). No relevant movements in the chamber nor extreme behavior reactions (Gonçalves et al., 2012) were recorded in any video sampling.

Tissue ṀO₂ was measured in 0.20 μm filtered seawater containing an additional 10 mM KCl, 200 mM taurine, and 1 mM glucose. The concentration of taurine used here is representative of measured tissue taurine levels in this species (Maccormack et al., 2016). Cuttlefish were euthanized as described above and 5–10 mg samples of gill, systemic heart, or mantle muscle were collected, weighed, and cut into 2–3 mg pieces using a razor blade. Tissue preparations from the same animal were then transferred to paired respirometry chambers (OX1LP, Qubit Systems Inc., Kingston, ON, Canada) containing 750 μL incubation medium maintained at 20 ± 0.1°C with a recirculating refrigerated water bath. Chambers were calibrated daily and background O₂ consumption rates were negligible without tissue. Baseline ṀO₂ was recorded for 300 s before addition of pharmacological agents and recordings were continued for an additional 300 s before experiments were terminated. Treatments consisted of 5 mM ouabain (e.g., Postel et al. (2000)) or 25 μM cycloheximide, a known antibiotic inhibitor of protein synthesis (Giuditta et al., 1968; Prozzo and Giuditta, 1973; Wieser and Krumschnabel, 2001) that has previously been applied in studies associated with learning in the species by blocking protein synthesis (Agin et al., 2003). Stock solutions were prepared in DMSO and delivered to the chambers in 10 μL. A tissue preparation from the same animal was run simultaneously with each drug treatment and exposed to 10 μL DMSO as a vehicle control. Chambers were thoroughly rinsed with 95% ethanol, ddH₂O, and incubation medium between runs and treatments were alternated between respirometry chambers to control for potential chamber effects. Respirometry chambers were interfaced to a LabQuest data acquisition system and DO₂ readings were collected over 10 min using LoggerPro V3.8 software (Vernier Software and Technology, Beaverton, OR, USA). Initial tissue ṀO₂ prior to treatment was determined from the linear decrease in DO₂ between 120 and 300 s and post-treatment ṀO₂ was determined between 400 and 600 s. ṀO₂ was linear with tissue mass (data not shown) and consistent with previous experiments (Maccormack et al., 2016).

All values are expressed as mean ± s.e.m with the exception of water temperature, salinity, and DO₂ that are expressed as mean ± standard deviation. Statistical tests applied are stated in either the Results section or the legends to the figures. All data were tested for normal distribution with the Shapiro-Wilk test as well as for homogeneity of the variances with the Levene's test (Zar, 1999). Statistical difference was considered for P < 0.05.

Figures 1A,B show examples of dissolved oxygen variation over the sampling period for individual cuttlefish either exposed to 100 or 50% DO₂ over 90 or 60 min, respectively. Oxygen consumption by hypoxic animals was 37% lower than animals held at 100% air saturation (189 ± 23 vs. 119 ± 6 nmol/g min) (Figure 1C). Cuttlefish maintained under control normoxic conditions exhibited a decrease in ventilation rate over the 60 min holding period (y = −0.52x + 88) (Figure 1D). Animals held under hypoxia also displayed a decrease in ventilation rate but to a lesser extent (y = −0.1x + 102). After 60 min ventilation rate was 85% higher in hypoxic than normoxic individuals.

Octopine was significantly higher in mantle than in gill or heart regardless of oxygenation condition. Octopine levels were significantly higher in mantle of hypoxic than normoxic animals but there was no significant difference between levels in normoxic or hypoxic animals in either gill or heart. (Figure 2A). There was no significant difference in either free glucose or glycogen levels in mantle between normoxic or hypoxic held animals (Figure 2B). HSP 70 was significantly higher in gill than in mantle or heart regardless of oxygenation condition (Figure 2C). There was no significant difference between levels in normoxic or hypoxic animals in any of the three tissues. Similarly, there was no significant difference in levels of polyubiquinated protein in mantle, gill or heart between treatment groups (Figure 2D).

Pre-treatment oxygen consumption rates for gill, heart, and mantle are presented in Figure 3A. There was no significant difference in rates amongst the three tissues with values for all being approximately 500 nmol O₂/gm wet weight. min. Rates for gill and heart are from animals weighing approximately 59 g; while those for mantle are from smaller/younger animals with a mass of approximately 2 g. Preparations were treated with cycloheximide and ouabain to assess how much of the decrease in oxygen consumption noted in the whole animal study could potentially be attributed to protein synthesis and Na/+K⁺ ATPase, respectively. Figure 3B shows the percentage decrease in oxygen consumption for each of the tissues following correction for any DMSO effect. Cycloheximide had a substantial effect upon oxygen consumption with average decreases ranging from 32% for gill to 42% for mantle. Similarly, ouabain treatment resulted in average decreases in oxygen consumption from 24% for mantle to 54% for heart.

Cuttlefish remained sedentary in the respirometry chambers under both normoxic and hypoxic conditions. As such, any difference in oxygen consumption between groups cannot be attributed to spontaneous swimming activity. When corrected for differences in temperature and body mass, the ṀO₂ values reported here under control conditions are in the same range as rates previously noted (Johansen et al., 1982; Melzner et al., 2006, 2007; Grigoriou and Richardson, 2009; Lamarre et al., 2016). Here, exposure to 50% DO₂ for 60 min resulted in a 37% decrease in oxygen consumption. Although limited to 2 individuals of S. officinalis with a mean mass of 153 g, a decrease in ṀO₂ of about 29% at a similar level of hypoxia was previously observed by De Wachter et al. (1988). Under comparable conditions, S. officinalis of body mass between 100 and 1,500 g showed a decrease in oxygen consumption of approximately 50% (Johansen et al., 1982). It is clear that a substantial reduction in ṀO₂ is a common response to modest hypoxia in S. officinalis; whether life could be sustained indefinitely at this level of hypoxia and if this will impact on the welfare (pain, suffering, distress, and lasting harm) of the animal remains unknown.

Ventilation rate under normoxic conditions decreased over the 60 min holding period in the same fashion as previously shown (Boal and Ni, 1996) and is presumably related to the stress of transferring the animals to the experimental chamber. After 60 min the ventilation rate recorded here was in the same range as previously noted in other studies (Bone et al., 1994; Boal and Ni, 1996; Boal and Golden, 1999; Melzner et al., 2006). Ventilation rate in hypoxic animals remained relatively stable with time and was 85% higher in hypoxic than normoxic cuttlefish after 60 min. To our knowledge this is the first report of rate of ventilation in hypoxic S. officinalis or any other cuttlefish species. The response differs from inactive, juvenile Humbolt squid where deep hypoxia resulted in a decrease in ventilation frequency, as well as contraction strength and ventilation volume per min (Trübenbach et al., 2013a). The mantle muscle of squid and cuttlefish consists of a main mass of circular fibers that are mitochondria-poor and primarily anaerobic, as well as inner and outer layers of mitochondria-rich aerobic fibers (Bone et al., 1981; Mommsen et al., 1981). The central fibers are used in escape jetting contractions while the fibers of the inner and outer layers are responsible for rhythmical respiratory ventilation (Bone et al., 1981). It is probable that the increase in ventilation frequency observed here places additional demand to supply ATP to the contractile mechanisms under hypoxia even in the face of reduced oxygen consumption.

Octopine dehydrogenase is the dominant opine dehydrogenase in S. officinalis. We were unable to detect activity of alanopine, tauropine, or strombine dehydrogenase in this species' tissues (unpublished data). The level of ODH in mantle, heart and gill was 35, 20, and 2 μmol/min g wet weight, respectively (Speers-Roesch et al., 2016). The level of octopine in mantle was substantially higher than in heart or gill, consistent with the high activity of ODH in mantle and lower activity in the other two tissues.

Sixty min of 50% hypoxia was associated with an increase in mantle octopine of 2.5 μmol/g. There was no change in octopine level in either gill or heart following the hypoxic period. The observed increase in octopine in mantle was low relative to the capacity to produce this metabolite. For instance, the increase in mantle octopine in S. officinalis forced to exercise and then subjected to hypoxia was 13 μmol/g (Storey and Storey, 1979). Another study found a drastic elevation of octopine in the haemolymph of S. officinalis exposed to <30% O₂ hypoxia for about 60 min (Storey et al., 1979). A similar increase in mantle octopine occurred in hypoxic Humbolt squid (Seibel et al., 2014) and squid (Lolliguncula brevis) forced to swim to exhaustion (Finke et al., 1996). The modest increase in octopine observed in this study is in line with the lack of change in glycogen levels during the hypoxic challenge. Based on ATP equivalents that may be generated by aerobic and anaerobic metabolism, the contribution of octopine production to overall energy production is minimal.

The relative quantity of polyubiqutinated protein was similar in mantle, gill, and heart under control conditions. There was no impact of the hypoxic challenge in any of the tissues tested. This finding is in contrast to that for juvenile Humbolt squid where severe hypoxia resulted in a three-fold increase in ubiquitinated proteins with a size of approximately 100 kDa (Trübenbach et al., 2014). These authors proposed that the increase in polyubiquintinated protein in association with a decrease in levels of at least two abundant proteins in the same size range (HSP90 and alpha-actinin) provides protection under hypoxia via a number of mechanisms, including the provision of certain amino acids for anaerobic energy production (Trübenbach et al., 2014). In S. officinalis food deprivation results in almost total depletion of digestive gland triglyceride, ṀO₂ decreases by approximately 30%, and mantle polyubiquitin mRNA and polyubiquitinated proteins increase, suggesting an increase in protein degradation via the ubiquitin-proteosome pathway to provide amino acids for energetic purposes (Lamarre et al., 2012, 2016). Although in this species there appear to be mechanisms to increase protein breakdown in association with decreased ṀO₂, we found no evidence that this occurs under modest hypoxia, at least based on levels of polyubiquinated protein.

HSP70 is a well-recognized stress protein that plays a role in maintaining protein integrity via chaperoning newly synthesized and unfolding proteins (Balchin et al., 2016). HSP70 levels were significantly higher in gill than in mantle or heart and did not change in any tissue with exposure to hypoxia. Relatively high HSP70 levels in S. officinalis gill may be related to the high rate of protein synthesis in gill relative to mantle based on direct measurements (Lamarre et al., 2016), and relative to heart and mantle based on total RNA levels (Lamarre et al., 2012). Furthermore, since gill is at the interface with the aquatic environment it may experience more perturbations than other tissues and so has higher constitutive levels of HSPs as protection. HSP70 levels increased in mantle of hypoxic juvenile Humbolt squid (Trübenbach et al., 2013b) but not in adult animals (Seibel et al., 2014). Thus, the generality of a HSP70 response to hypoxia in cephalopods is unresolved but it is clear that under the conditions of the current study there was no change in HSP70 level.

The minimal increase in octopine content, along with stability in glycogen pools, polyubiquitinated protein level, and HSP70 levels in cuttlefish subject to 50% hypoxia for 1 h suggests that anaerobic metabolism is activated to only a limited extent and that the challenge is not particularly stressful to the animal. The 37% decrease in ṀO₂ must be associated with an equivalent decrease in energy demand. Given that the animals are quiescent under both control and hypoxic conditions, the question becomes as to what metabolic processes are decreased to maintain energy balance.

There was no difference in the rate of oxygen consumption between tissue slices of gill, heart, and mantle. The in vitro activity of citrate synthase, a qualitative indicator of aerobic metabolism, was approximately 20-fold higher in S. officinalis heart than in mantle or gill which were similar to one another (Speers-Roesch et al., 2016). The relatively low rate of oxygen consumption in heart noted here is likely due to the non-contractile nature of the preparation and reflects the basal metabolism of quiescent cells. The heart and gill samples were from juvenile animals and the mantle from hatchlings. Smaller animals have higher mass specific rates of oxygen consumption than larger animals (Johansen et al., 1982; Grigoriou and Richardson, 2009). It is likely that if oxygen consumption of mantle had been conducted on tissue removed from animals the same size as that used for gill and heart that the rate for mantle would have been relatively lower.

One of the major energy demanding processes in non-contracting cells is protein synthesis. In food deprived cuttlefish, both ṀO₂ and protein synthesis are decreased in concert. A 36% decrease in whole animal ṀO₂ was associated with a 63% decrease in fractional rate of protein synthesis in both mantle and gill (Lamarre et al., 2016). The importance of these findings in the context of the current study is that it demonstrates that protein synthesis is a regulatable element of the energy demand machinery in S. officinalis. Here we show that the inclusion of cycloheximide decreases ṀO₂ in isolated gill, heart, and mantle preparations. Based on studies with goldfish and trout it is likely that the level of cycloheximide utilized here resulted in a total shut down of protein synthesis (Wieser and Krumschnabel, 2001), as previously suggest by the Agin et al. (2003) results on the species and because the amount used here was more than 3 times higher than the dosages used in Octopus vulgaris (Prozzo and Giuditta, 1973) and Loligo pealii (Giuditta et al., 1968), which resulted in 80% protein synthesis inhibition in the optic globe with 100 μg/mL and 95–96% inhibition in squid axons with 200 μg/mL, respectively. If it is assumed that all tissues in the quiescent animals are consuming oxygen at similar rates, a total inhibition of all protein synthesis would be required to meet the whole animal decrease in ṀO₂ of 37%. This is highly unlikely given that even in an anoxic resistant fish, oscar (Astronotus crassipinnis), protein synthesis is decreased by only 55% and 60–85% in muscle and liver, respectively, at a DO₂ of 10% (Lewis et al., 2007). Therefore, although a decrease in protein synthesis could potentially contribute to the decrease in whole animal ṀO₂, there must be additional mechanisms to reduce energy expenditure.

Na⁺/K⁺ ATPase is a major driving force for many energy dependent ion transport processes and is considered one of the major cellular sites for ATP utilization.

Na⁺ /K⁺ ATPase has been identified in numerous tissues of S. officinalis (Donaubauer, 1981) and is presumably ubiquitous. A ouabain sensitive Na⁺/K⁺ ATPase has been identified in gill of squid (Doritheutis plei) (Proverbio et al., 1988) and more recently, Na⁺/K⁺ ATPase in Sepia gill was shown to increase in response to increases in water CO₂ level (Hu et al., 2011). In the hypoxia tolerant intertidal clam, Mercenaria mercenaria, a severe hypoxic exposure resulted in a decrease in the maximal in vitro enzyme activity of Na⁺/K⁺ ATPase (Ivanina et al., 2016). As with protein synthesis, these studies illustrate that Na⁺/K⁺ ATPase is a site that can be controlled to alter energy demand. Here it is shown with isolated tissues that inhibition of Na⁺/K⁺ ATPase with ouabain decreased ṀO₂ by ~20% in mantle and 50–60% in gill and heart. Given that the Na⁺/K⁺ ATPase is one of the largest consumers of ATP in cells, it is likely that a curtailment of this process occurs in S. officinalis under hypoxia.

The handling Editor declared a past co-authorship with one of the authors AS and states that the process nevertheless met the standards of a fair and objective review. The other authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.