SUMMARY
Hooded seals (Cystophora cristata) rely on large stores of oxygen, either bound to hemoglobin or myoglobin (Mb), to support prolonged diving activity. Pups are born with fully developed hemoglobin stores, but their Mb levels are only 25–30% of adult levels. We measured changes in muscle [Mb] from birth to 1 year of age in two groups of captive hooded seal pups, one being maintained in a seawater pool and one on land during the first 2 months. All pups fasted during the first month, but were fed from then on. The [Mb] of the swimming muscle musculus longissimus dorsi (LD) doubled during the month of fasting in the pool group. These animals had significantly higher levels and a more rapid rise in LD [Mb] than those kept on land. The [Mb] of the shoulder muscle, m. supraspinatus, which is less active in both swimming and hauled-out animals, was consistently lower than in the LD and did not differ between groups. This suggests that a major part of the postnatal rise in LD [Mb] is triggered by (swimming) activity, and this coincides with the previously reported rapid early development of diving capacity in wild hooded seal pups. Liver iron concentration, as determined from another 25 hooded seals of various ages, was almost 10 times higher in young pups (1–34 days) than in yearling animals and adults, and liver iron content of pups dropped during the first month, implying that liver iron stores support the rapid initial rise in [Mb].
INTRODUCTION
Seals and whales depend on oxygen stores bound to hemoglobin (Hb) in the blood and to myoglobin (Mb) in skeletal muscles to support aerobic metabolism during diving, and their dive capacity is, accordingly, positively correlated with blood volume, blood [Hb] and muscle [Mb] (Mottishaw et al., 1999; Noren and Williams, 2000).
Adult hooded seals (Cystophora cristata) have the largest oxygen-storing capacity per unit body mass of any mammal hitherto examined (Burns et al., 2007; Lestyk et al., 2009), and may stay submerged for more than 1 h (Folkow and Blix, 1999). Newborn hooded seals, in contrast, have a mass-specific blood oxygen-storing capacity that is similar to that of adults, but their muscle [Mb] is only about 25% of the adult value (Burns et al., 2007). Even so, these pups, which have the shortest lactation period of any mammal (Bowen et al., 1985), also have a very rapid development of diving capacity, and may dive to depths of >100 m for durations of >15 min before reaching 3 weeks of age (Folkow et al., 2010).
The development of oxygen storage capacity is known to involve hypoxia-inducible factor-1 (HIF-1), which among other things regulates the expression of erythropoietin and hence synthesis of Hb (Gassmann and Wenger, 1997). The seal fetus is repeatedly exposed to hypoxia in utero, when the mother is diving (Elsner et al., 1969; Liggins et al., 1980), and this most likely triggers the expression of HIF-1 and the development of blood O2 stores before birth. Mb expression, however, has been shown to be regulated by the calcium–calcineurin–NFAT (nuclear factor of activated T-cells) pathway (Bassel-Duby et al., 1993; Kanatous et al., 2009), which in turn is also partly triggered by hypoxia, but apparently only in combination with muscular activity, at least in mice (Kanatous et al., 2009; Wittenberg, 2009). A recent study using cultured myocytes from Weddell seals (Leptonychotes weddellii) showed that control of Mb development in these animals differs somewhat from that outlined above, in that Mb levels were found to increase in response to hypoxia (and also after lipid supplementation), even in the absence of contraction (De Miranda et al., 2012). The authors nevertheless concluded that a secondary stimulus would still be required to increase Mb levels to those seen in the whole animal. As movement is rather restricted for the fetus, muscular activity may be the trigger that is missing for full Mb development in utero in the hooded seal.
For Mb development to occur, some source of iron would be required, given this element is a crucial component of this and other heme proteins. As hooded seal pups may undergo an extended fasting period after their very brief nursing period (e.g. Bowen et al., 1987; Oftedal et al., 1989), any Mb synthesis in this period must be based on endogenous stores of iron. The major compartments for iron storage in mammals are the Hb pool and the liver (e.g. Fleming and Bacon, 2005; Graham et al., 2007). As it would be counter-productive to catabolize Hb in order to obtain iron for Mb synthesis, we hypothesized that any Mb development that takes place during fasting in these pups is likely to primarily be supported by mobilization of ferritin-bound iron from their liver.
The present study, thus, had a dual purpose: first, to describe the early postnatal development of the myoglobin stores in hooded seal pups and investigate whether muscular activity affects this development; and second, to determine whether their liver is a likely source of iron for postnatal Mb development.
MATERIALS AND METHODS
Animals
A total of 33 hooded seals, C. cristata (Erxleben 1777), were used in this study. Eight newly weaned pups (aged 4–6 days, as judged from external characteristics and the presence/absence of a mother) (Bowen et al., 1987) were collected in the pack ice of the Greenland Sea (the West-ice, at about 71°40′N, 14°20′W) between 22 and 27 March 2010 during a research cruise with R/V Jan Mayen, under permits from the Norwegian and Greenland authorities. These pups were brought to Tromsø, Norway, where they were kept in the approved research animal facility of the Department of Arctic and Marine Biology at the University of Tromsø, for studies of postnatal changes in muscle Mb levels. The animals were divided into two groups: four animals were kept in a 40,000 l indoor seawater pool that permitted swimming activity (pool group), while the remaining four pups were kept in a fenced snow-covered ~25 m2 outdoor enclosure with no possibility of swimming (land group). After ~2 months (age of pups ~70 days), when conditions no longer allowed the latter to be maintained outdoors because of snow-melt, the land group was transferred indoors to another 40,000 l seawater pool, which was identical and adjacent to the one already occupied by the pool group. In both pools, a wooden ledge allowed the seals to haul out as desired. To mirror the naturally occurring 1 month post-weaning fast (Bowen et al., 1987), captive pups were not fed until the age of ~37 days, after which they were fed capelin (Mallotus villosus) and subsequently herring (Clupea harengus). Dietary intake of fish gradually increased from a few fish per day to 2.0–2.5 kg of fish per day by the time pups were ~2 months old and onwards. The diet was supplemented with a vitamin complex (Sea Tabs II for marine mammals, Pacific Research Laboratories, CA, USA). All animals were subjected to normal light/dark cycles at 69°N latitude, either natural (outdoor land group) or simulated (indoor pool group).
Twenty-one other hooded seals (12 pups, aged 1–10 days; four yearlings; and five adults, aged 2–16 years) were culled in the same general area (~72°00′N, 16°40′W) in which the eight pups were caught, during similar research cruises to the Greenland Sea in March 2011 and 2012, for collection of miscellaneous data and tissue samples for a range of scientific purposes, including liver mass and samples for analyses of liver iron contents for the present study. Additional liver mass data were obtained from our own unpublished records (L.P.F., unpublished data) and from published data (Oftedal et al., 1989). The pups were aged from external characteristics and the presence/absence of a mother (Bowen et al., 1987), yearlings from external characteristics, and adults based on inspection of sectioned teeth under the microscope as described elsewhere (Rasmussen, 1960).
Four newly weaned pups were captured during the 2011 research cruise to the Greenland Sea and transferred to Tromsø where they were kept under conditions that were identical with those of the pool group in 2010, i.e. including a 30 day fasting period (see above). These pups were killed at the age of ~34 days for collection of various samples, including liver samples for analysis of iron content.
All animals were killed in accordance with the Norwegian Animal Welfare Act, through stunning [shot through the head or given an overdose of pentobarbital (Nembutal, 20 mg kg−1) injected into the extradural intravertebral vein] immediately followed by bleeding through severing of the brachial vasculature. All use of research animals was approved by the National Animal Research Authority of Norway (permit no. 2402).
Muscle, blood and liver sampling
Muscle biopsies were collected at intervals from the main swimming muscle (musculus longissimus dorsi, LD) and from one of the flipper muscles (musculus supraspinatus, SSP) of the eight pups that were captured in 2010. The pups were first sampled when about 1 week old, while the last biopsies were collected when they were ~1 year of age (394 days). The sampling interval during the first month was 10 days, which was then extended to every 2 weeks during the second month, and was bi-monthly from there on. Prior to sampling, animals were weighed (model 235 suspended weight; Salter, Tonbridge, Kent, UK) and then sedated with an intramuscular injection of Zoletil Forte Vet (1–1.5 mg kg−1; tiletamin–zolazepam, Virbac, Carros Cedex, France) supplemented when needed with i.v. injections (0.2–0.3 mg kg−1) via a venous catheter (Secalon T, IGG/1.7×160 mm; Becton Dickinson, Franklin Lakes, NJ, USA) in the extradural intravertebral vein. Muscle biopsies were collected under additional local anesthesia, after injection of Xylocaine (3–4 ml of 10 mg ml−1; Astra Zeneca, Södertälje, Sweden) in the incision region. The skin area (~3×3 cm) was shaved and disinfected with chlorhexidine (5 mg ml−1; Fresenius Kabi, Halden, Norway). A small incision was then made with a sterile scalpel (blade no. 11) and the sample was collected using a sterile disposable 6 mm biopsy punch (Miltex, York, PA, USA) that was advanced through the skin and blubber into the underlying muscle. Collected (duplicate) samples were temporarily stored on ice and then transferred to cryovials and frozen at −80°C for later analysis. After sampling, the incision site was closed with absorbable suture and the animals returned to their holding facilities immediately after recovery from sedation. Sites for sampling were alternated between the two body sides, with LD sampling sites being located in the upper lumbar region. On each sampling occasion, a blood sample was also collected via the central venous catheter in the extradural intravertebral vein, for determination of hematocrit (Hct) values by use of a hematocrit centrifuge (EBA 12, Type 1000, Hettich, Tuttlingen, Germany).
Liver samples were collected from the 25 culled animals of various ages (Table 1), by cutting a central piece (~4 cm3) from the biggest of the six liver lobes. The samples were frozen at −20°C for later analysis of iron content.
Mb analyses
[Mb] was determined according to Reynafarje (Reynafarje, 1963), as previously described (Burns et al., 2007; Lestyk et al., 2009). Briefly, frozen muscle samples were thawed, cleaned of connective tissue and blood and sonicated (Sonic Dismembrator model 500, Fisher Scientific, Fairlawn, NJ, USA) in ice-cold 0.04 mol l−1 phosphate buffer (19.25 ml g−1 tissue, pH 6.6). The samples were then centrifuged at 10,000 g for 5 min at 4°C. The supernatant was placed in a vacuum chamber, which was first gassed with CO (99.5%) for 30 s, then filled with CO for 15 s and closed. After 20 min of CO incubation, sodium dithionite solution was added (1% in sample) and the samples were vortexed to ensure full reduction of Mb for correct absorption measurement. After an additional 5 min of CO incubation, the optical density was read at 538 and 568 nm (Spectra Max 340PC, Molecular Devices, Sunnyvale, CA, USA) and [Mb] was calculated.
Assays were run in triplicate and each run included both lyophilized Mb (Sigma, St Louis, MO, USA) and tissue controls from an adult harbor seal (Phoca vitulina) with known Mb levels (Burns et al., 2007), to validate the results. To estimate the variance within the muscles, three large samples of LD and SSP muscles, respectively, were collected from one of the culled animals and analyzed. The precision of the assay was estimated from the harbor seal tissue controls.
Liver iron analyses
Non-heme liver iron content was determined as described previously (Rebouche et al., 2004), as non-heme iron concentration per tissue wet mass. Briefly, liver samples were thawed and homogenized in high-purity water (1:10 w/v) and protein precipitation solution [1 mol l−1 HCl and 10% trichloroacetic acid (Fisher Scientific) in high-purity water] was added (1:1). After incubation for 1 h at 95°C, the samples were vortexed and centrifuged at 8200 g for 10 min at room temperature. The supernatant was mixed (1:1) with chromagen solution [0.508 mmol l−1 ferrozine (Sigma), 1.5 mol l−1 sodium acetate (Fisher Scientific) and 0.1% or 1.5% (v/v) thioglycolic acid (Sigma) in high-purity water]. After 30 min incubation at room temperature, the samples were centrifuged at 8200 g and the optical density of the supernatant (triplicates) read at 562 nm (Spectra Max 340PC). An iron standard (Sigma) and sample blanks were prepared by mixing the supernatant with chromagen solution without ferrozine. Liver water content was determined by drying separate samples from each age class for 48 h at 70°C. Data on liver masses were used to estimate liver iron content.
Data handling and statistics
A linear mixed model approach was used to analyse the Mb data. To determine the effect of muscle type, group location and age on Mb, each of these parameters was tested in a full model that included other parameters as fixed factors. If the muscle type had a significant effect, i.e. there was a significant difference between the muscle types, the data were split and analysed for the two types separately. If group location had a significant effect, the data were further split and analysed for each group, to determine the effect of time. In addition, a pair-wise comparison of the different age classes was conducted. If a parameter had no significant effect, it was removed from the model. To determine the effect of group location for each age class (i.e. at each sampling occasion), the data were split to analyse each separately. Between-groups comparisons were only made until age ~70 days, as after that time both the pool and the land groups were maintained in (separate) pools. Pairwise comparisons were adjusted for multiple comparisons with the Bonferroni method.
Age-wise comparisons of liver iron concentration and content were made using independent samples t-tests.
For all tests, P<0.05 was considered significant. Values are presented as means ± s.d. unless otherwise stated. All statistical analyses were made using SPSS v.19.0 (SPSS Inc., Chicago, IL, USA).
RESULTS
The body mass of the eight pups that were kept in Tromsø for collection of muscle biopsies dropped by ~30% from 44.2±3.8 kg to 32.3±2.2 kg during the 30 day fasting period and thereafter increased to 68–85 kg at 1 year of age (Fig. 1A). There was no significant difference in body mass between the two groups (F1,70=1.250, P=0.267). At the age of ~70 days, the land group was reduced from four to three animals, as a result of the death of one pup during sedation.
Hct values were similar in the two groups and remained stable at around 60% (overall mean Hctpool=58.87±4.38%, Hctland=59.29±4.35%) throughout the period (Fig. 1B).
There was a significant difference in LD [Mb] between the pool and land group (F1,35=10.9, P=0.002, mean difference pool–land=5.91±1.79 mg Mb g−1), with pool animals having higher levels than land animals (Fig. 2). For both groups, LD [Mb] showed a significant increase with age (pool: F5,18=6.878, P=0.001; land: F5,17=4.912, P=0.006), until 65 days, but then leveled off (Fig. 2).
With regard to SSP, there was no statistically significant difference in Mb levels between the two groups. Moreover, age had a significant influence on the SSP Mb values only in the pool group (F6,17=4.077, P=0.013). Mean Mb levels were significantly lower in SSP than in LD in both groups and at all ages (pool: F1,35=23.883, P<0.001; land: F1,34=4.434, P=0.043), but the mean difference was higher in the pool group than in the land group (pool: LD–SSP=7.95±1.63 mg Mb g−1; land: LD–SSP=3.58±1.70 mg Mb g−1. The SSP Mb levels, like the LD Mb levels, rose steadily at the beginning of the sampling period, but then leveled off (estimated variation within the muscles: CVLD=5.7%, CVSSP=6.1%; estimated precision of the assays: 95%).
Data on body mass, liver mass and liver iron concentration of the seals that were used for liver iron content analyses are shown in Table 1. Liver mass on average corresponded to 2.22% of adult body mass and 2.18% of yearling body mass, which was lower than that in young pups (2.92%), but higher than that in the fasted pups (1.59%) (Oftedal et al., 1989). There was no significant difference in liver mass within the young pup age class even though weaned pups tended to have slightly higher values (newborn: 843±453 g, N=10; weaned: 1114±433 g, N=9).
Liver iron concentration displayed large variance, even within age classes. There were no significant differences in liver iron concentration among pups aged 1–34 days (F=0.043, P=0.838), although fasted pups had somewhat higher concentrations than young pups (Table 1). There were no significant differences in liver iron concentration among older (>1 year) animals (F=0.152, P=0.708). Overall, pups (1–34 days) had significantly higher liver iron concentrations than older animals (986±508 versus 140±89 μg g−1 wet mass, F=19.69, P<0.001) (Fig. 3).
The estimated mean total liver iron content was highest immediately after birth, in both absolute and relative (mass-specific) terms (Table 1). During the fasting period (during which LD Mb levels doubled), liver iron content decreased by 25%, but was still higher than in yearlings, which also had had access to dietary iron to sustain development of Hb and Mb levels during growth.
The liver water content ranged between 60.0 and 63.6% in newborns (1–10 days) and 69.1 and 71.6% in the older animals (34 days to 16 years).
DISCUSSION
Blood Hct was stable at the adult level (~60%) throughout this study (Fig. 1B), which confirms that hooded seals have fully developed blood oxygen stores from the very beginning of life (Burns et al., 2007).
However, we have shown for the first time that while [Mb] is low at birth, there is a rapid initial increase in [Mb] in the LD muscle of hooded seal pups, from 25% to 50% of adult values, within their first month of life (Fig. 2). It is intriguing that this remarkable rise in [Mb] coincides closely with the development of diving capacity in these animals (Fig. 4) (Folkow et al., 2010).
We further conclude that this postnatal rapid rise in [Mb] was mainly due to increased muscular activity. We base this conclusion on the following arguments. The pool group, which indulged in vigorous swimming, known to involve the LD muscles (Kanatous et al., 1999), both showed a more rapid development of LD [Mb] and maintained significantly higher [Mb] at all times than did the LD muscle of animals in the land group (Fig. 2), which typically were immobile for most of the time and certainly did not engage in swimming or other activities that involved the LD muscles. We further note that [Mb] in the SSP muscles, which apparently were not used to any particular extent either in the pool or in the land group, was both significantly lower than in the LD muscle (regardless of group) and quite similar between the two groups.
If we, for the sake of the argument, assume that (diving-induced) hypoxia, instead of muscular activity, is the main trigger for Mb development, we would expect between-groups differences in Mb levels for both muscle types (i.e. also for the SSP muscles), but this was not the case. It could be argued that the dives of the pool group animals were not of sufficient duration (typically 1–2 min) to produce any substantial hypoxia, and that land group animals might have experienced equivalent hypoxia in connection with sleep apnea, but in that case there is no good reason why LD Mb levels should differ between the groups – unless Mb development requires muscular activity, as we have concluded. Thus, the mere fact that [Mb] increased at a similar rate in the SSP but at a different rate in the LD in the two groups suggests that activity, at least, is a crucial component for the development of Mb, which is in agreement with findings in other species (Pattengale and Holloszy, 1967; Kanatous et al., 2009; Wittenberg, 2009).
Our captive animals had lower [Mb] at 1 year of age than wild animals (Burns et al., 2007), which is probably due to reduced activity and a much lower exposure to hypoxia in captivity. Similar observations and conclusions have been made for other diving species (e.g. emperor penguins, Aptenodytes forsteri) (Ponganis et al., 2010). This implies that wild animals probably show an even more remarkable initial rise in [Mb] than our captive pups.
It is worth noting that a significant part of the rapid postnatal rise in [Mb] occurred while our pups were fasting, which happens naturally at that time (Bowen et al., 1987). As blood Hct values remained stable, we also assume that no catabolism of blood Hb took place that could make iron available for Mb synthesis. We hypothesize that the iron required for this rapid production of Mb must have been derived from other internal stores, such as the liver. This organ is known to represent a key compartment for storage of iron in humans and other species, in which dietary iron is taken up from the portal blood, and, at times of increased demand, released into the circulation (e.g. Fleming and Bacon, 2005; Graham et al., 2007). This hypothesis is supported by our observation of significantly higher liver iron concentrations (and content) in pups than in older animals (Fig. 3). Also, although fasted pups apparently had higher liver iron concentrations than did newborn pups (but this difference was non-significant), their liver mass was only about half that of the newborns, which means that liver iron content decreased substantially during the 1 month of fasting (Table 1). The reason why the hooded seal liver undergoes this large mass change during fasting is not known, but it probably relates to mobilization of large liver-based glycogen stores during the early part of the fast, as previously shown in other fasting/starving mammals (e.g. Nilsson and Hultman, 1973).
If we assume that blubber was the main source of energy during the postnatal fasting period and that muscle mass thereby remained largely unchanged, as in grey seal (Halichoerus grypus) pups (Nordøy and Blix, 1985), the amount of Mb produced may be estimated based on data from Burns and associates (Burns et al., 2007), which show that the skeletal muscle mass of a weanling hooded seal corresponds to about 20% of its body mass, or 8.8 kg in a 44.2 kg pup (which was the average mass of our eight weanlings; Fig. 1A). With an average increase in [Mb] in this muscle mass from ~25 to ~35 mg g−1 within the fasting period (Fig. 4B), about 88 g of Mb may have been synthesized, which would require mobilization of 280 mg of iron (given the need for one Fe for each Mb molecule and a molecular weight ratio for Fe/Mb of 55.84/17,380). The estimated decrease in liver iron content in the same period was about 182 mg (Table 1), which is fairly close to the estimated iron required for Mb synthesis, particularly when considering the remarkable variation in the liver iron concentration in seals, as in other species (Zuyderhoudt et al., 1978; Faa et al., 1994; Bonkovsky et al., 1999). In estimating liver iron content in young pups (newborn/weaned; Table 1), we have pooled the data on liver mass and liver iron concentration for all pups aged 1–10 days, as there was no significant difference in the liver data within this age group. Moreover, hooded seal pups are physically inactive during the suckling period, which is largely spent suckling or sleeping. Although milk-derived lipids may have stimulated some Mb development (Burns et al., 2010; De Miranda et al., 2012), we do not expect that much liver iron was mobilized for Mb synthesis during lactation, as this period lasts for only about 4 days in this species (Bowen et al., 1987).
It may appear puzzling that Mb, unlike Hb, is not well developed from birth in these soon-to-become expert diving seals, as fully developed Mb stores would allow a more rapid development of diving capacity. However, pregnant seal mothers have been shown to continue their normal diving pattern until shortly before delivery (Liggins et al., 1980) and it has been shown indirectly that the fetus, unlike the mother, depends on chemoreceptor stimulation to develop the cardiovascular defense responses to hypoxia (Elsner et al., 1969). This implies that the instant selective vasoconstriction and bradycardia that are characteristic of adult animals in dives of long duration (Blix and Folkow, 1983; Ramirez et al., 2007) take time to develop in the fetus. Moreover, as Mb has a much higher affinity for oxygen than does Hb, a high concentration of Mb would, with no benefit to the immobile fetus, deplete the blood of oxygen, particularly at the end of dives, when maternal arterial oxygen tension sometimes reaches extremely low levels. It therefore appears to be more advantageous to build Mb stores at a rapid rate after birth, as seems to be the case in hooded seal pups, rather than have high levels from birth.
Acknowledgements
We thank Professor E. S. Nordøy and the crew of R/V Jan Mayen for assistance in the field, and Dr Chandra Ravuri for laboratory assistance.
FOOTNOTES
FUNDING
S.J.G., A.S.B. and L.P.F. were supported by faculty grants from the Department of Arctic and Marine Biology, University of Tromsø, and J.M.B. was supported by a faculty grant from the College of Arts and Sciences, University of Alaska, Anchorage.
REFERENCES
COMPETING INTERESTS
No competing interests declared.