Introduction

In Nunavik (Canada), beluga mattaaq—an Inuit delicacy consisting of the skin and underlying fat—constitutes an important dietary source of the organoselenium compound selenoneine (SEN), particularly in the Hudson Strait region (Fig. 1), where beluga hunting is central to the local diet1. SEN and its sulphur isologue ergothioneine (EGT), closely related betaine derivatives of histidine, are key antioxidants and cytoprotectants2,3. While EGT may protect humans against several chronic diseases related to aging4, SEN may offer protection against methylmercury (MeHg) toxicity5, a contaminant of great concern for Nunavimmiut, for whom food insecurity is high and subsistence diets include some marine species with elevated MeHg levels6. First discovered in bluefin tuna7, SEN is abundant in the flesh of predatory marine species such as tuna and swordfish8. Until recently, EGT was thought to be limited mainly to terrestrial food sources such as mushrooms and grains4,9, but a recent report documented its presence along with SEN in tuna and swordfish10. This study focuses on beluga whales (Delphinapterus leucas), which hold significant cultural and dietary importance for Inuit populations and have been central to SEN research.

Understanding the distribution of these antioxidants in beluga tissues is needed to better assess the nutritional benefits of this species to local communities and may shed light on how belugas cope with Arctic environmental stressors. While Inuit communities consume various parts of the beluga, not only the skin but also meat, blubber, and intestines, the concentrations of SEN and EGT in these organs and tissues remain largely unexplored11,12. Notably, sources of EGT in the Inuit diet have yet to be identified despite a recent report highlighting high blood levels of EGT among Nunavimmiut, which appear to be driven by country food consumption13. Their structural similarity, the fact that they share a common cellular transporter—the ergothioneine transporter (ETT, coded by the human gene SLC22A4)5 —and the presence of microorganisms capable of synthesising both compounds in Arctic marine ecosystems14,15, led us to hypothesise that EGT could be present in marine mammals as well.

Through a close collaboration with Inuit hunters from Quaqtaq, a community located on the shores of the Hudson Strait, whose traditional knowledge and sustainable practices provided a unique access to otherwise unattainable samples, this study explores the distribution of key antioxidants in marine mammals. By way of our in-house chemical synthesis of standards and adaptation of our specific and accurate analytical method developed for blood to diverse tissues matrices, we were able to detect and quantify SEN, EGT, and their methylated metabolites Se-methyl-SEN, and S-methyl-EGT16 in skin, muscle, liver, kidneys, intestine, brain and blood of 14 adult belugas harvested between 2018 and 2019 (Fig. 1). We also investigated the potential transplacental transfer of SEN and EGT by analysing tissues from a beluga fetus. Additionally, we quantified these antioxidants across different skin layers and utilised immunofluorescence staining to localise ETT, further elucidating their distribution within this crucial storage tissue and key country food for Inuit. We discuss the potential role of SEN and EGT in protecting belugas from sun damage, thereby opening new avenues for research on skin cancer prevention, and against other stressors of the northern environment such as MeHg.

Fig. 1
Fig. 1The alternative text for this image may have been generated using AI.
Full size image

Map of Nunavik, highlighting the three main regions of eastern Hudson Bay (red), Ungava Bay (yellow) and Hudson Strait (blue), and the community of Quaqtaq. Map created using data from the Government of Canada with the Provinces/Territories Cartographic Boundary File (2016) and the Water File—Lakes and Rivers polygon (2013). Coordinate reference system: WGS84. Produced using ArcGIS Pro. Sampling information is also provided.

Inuit involvement in project and ethical sampling practices

This project relied on the expertise and essential contributions of local Inuit hunters in Quaqtaq, who provided samples in ethical agreement with the regional Anguvigaq and their local Anguviait, an Inuit association mandated to protect their harvesting rights while ensuring sustainability which provides direction on harvest, conservation and wildlife research in Nunavik. Beluga hunting is deeply rooted in Inuit tradition and plays a vital role in food security in the region11. This unique collaboration with Inuit enabled access to fresh samples from healthy, recently harvested animals—an invaluable opportunity for this study, as whale samples for scientific purposes often come from stranded animals, which may be unhealthy or decomposed. Moreover, the collection of a whale fetus from a pregnant female beluga presents an uncommon occurrence, and provided a rare opportunity to study placental transfer of these antioxidants. The Makivvik Corporation, Anguvigaq, Anguviait and the Nunavik Nutrition and Health Committee (NNHC) were all informed and agreed to the experimental protocols for the present work. Methods for sustainable hunting were carried out in accordance with local and regional guidelines, as well as Canadian federal regulations governing marine mammal harvesting for subsistence purposes17,18. This study did not involve any live animal experimentation; all samples were obtained post-mortem. The manuscript was submitted to the NNHC board for review and approval, accompanied by a plain language summary and infographics. Key findings were also presented on local radios and at the annual Anguvigaq meetings, which bring together hunters from all 14 Nunavik communities and are also broadcasted live on local radios and simultaneously translated into Inuktitut.

Results and discussion

SEN and EGT in beluga tissues

SEN and EGT concentrations in tissues of the 14 beluga whales harvested in 2018–2019 are presented in Fig. 2 (see Supplementary Table 1 for additional information on the harvested whales). Comparisons across tissues reveal that the skin is the primary reservoir for both compounds. Mean skin concentrations were 17.2 µg/g for SEN and 78.6 µg/g for EGT (wet mass). Internal tissues displayed statistically lower concentrations compared to the skin, with the highest mean concentrations found in the kidney (1.23 µg/g for SEN and 6.59 µg/g for EGT) and the brain (0.98 µg/g for SEN and 3.23 µg/g for EGT). These were followed by the intestine (0.52 µg/g for SEN and 2.68 µg/g for EGT) and whole blood (0.50 µg/g for SEN and 1.58 µg/g for EGT). Lower concentrations were observed in the liver (0.17 µg/g for SEN and 0.77 µg/g for EGT). The lowest levels were measured in the muscles (0.07 µg/g for SEN and 0.29 µg/g for EGT) (Fig. 2a, b).

SEN and EGT exhibited a similar distribution pattern across all tissues, with EGT concentrations consistently exceeding SEN concentrations. EGT:SEN molar concentration ratios were approximately 5:1 in most tissues, slightly higher in the kidney (6:1) and lower in the brain (4:1) and blood (3:1) (Supplementary Table 2). As sulphur is approximately 10⁵ times more abundant than selenium in humans under normal conditions26, the much lower EGT:SEN concentration ratios noted in beluga tissues suggest that their distribution is governed by compound-specific biological mechanisms rather than merely reflecting the availability of elemental precursors.

SEN distribution across internal tissues showed a trend similar to that recently reported in giant petrels by El Hanafi et al.19 who noted the highest concentrations in kidneys, followed in decreasing order by the brain, liver, blood and muscle. Higher concentrations were found in all internal organs of this marine bird (maximum SEN concentration in kidneys of 88 µg Se/g, dry weight, corresponding to 92 µg/g SEN, wet weight, based on a humidity of 70%) compared to belugas (maximum SEN concentrations of 3.26 µg/g, wet weight). Although skin was not studied for petrels, feathers showed low total selenium concentration20. High SEN (and EGT) concentrations found in beluga skin (mean 17.2 µg/g) might indicate an important functional role that directs SEN to the skin, resulting in lower levels in internal organs (Fig. 2a). These findings highlight the importance of considering skin concentrations in future studies investigating SEN and EGT distribution in animals.

Methylated metabolites of SEN and EGT, namely Se-methyl-SEN and S-methyl-EGT, were simultaneously quantified across all analysed tissues (Supplementary Fig. 1). Se-methyl-SEN was present in the skin at an average concentration of 41.9 ng/g, but was significantly lower in all other tissues (ranging from 2.17 to 4.65 ng/g), while S-methyl-EGT was measured in similar concentrations in all tissues (ranging from 7.46 to 29.8 ng/g) except the brain, in which it was not detected. Interestingly, prior to the present study, Se-methyl-SEN had been identified only in human blood and urine21,22. For instance, Se-methyl-SEN was not detected in any organs of the giant petrels or in our previous report on beluga skin19,21. This discrepancy could be attributed to methodological optimisations, including lower LODs and higher precision compared to previous studies. S-methyl-EGT, however, has been identified in human and animal tissues in the past, notably in mice and in the blood of human volunteers to whom EGT was administered23,24. In the latter studies, correlations between EGT and S-methyl-EGT were statistically significant for all tissues. Here, however, no correlations were observed amongst tissues between SEN and Se-methyl-SEN or between EGT and S-methyl-EGT. The lack of correlation between these antioxidants and their respective metabolites, as well as the high variation amongst individuals, suggests that they might not be methylated in belugas and that the metabolites might be acquired from their diet.

Fig. 2
Fig. 2The alternative text for this image may have been generated using AI.
Full size image

Mean concentration +/- standard error (SEM) of selenoneine (a) and ergothioneine (b) in µg/g (wet mass) in tissues of adult belugas hunted in Quaqtaq, Nunavik, 2018-2019. Sample sizes were n = 14 for skin, n = 13 for kidney and liver, n = 12 for muscle, n = 10 for blood and intestine and n = 5 for brain. Different letters identify statistically significant differences in mean concentrations across tissues (p < 0.05).

Our study is one of the few to report the presence of EGT in the marine environment, with one beluga skin sample containing as much as 182 µg EGT/g, the highest concentration ever recorded in a marine species. Prior to the present study, Fernández-Bautista et al.10 documented low levels of EGT in muscles of farm-raised salmon, tuna (Thunnus thynnus) (3.5 µg/g wet mass), and swordfish (1.1 µg/g wet mass). Ey et al.25 also reported a low EGT concentration in a single trout muscle sample (0.07 µg/g); the specific species was not mentioned and could belong to either freshwater or marine habitats. Their study also highlighted some of the highest EGT concentrations ever observed in terrestrial mushrooms, such as king bolete (Boletus edulis) and oyster mushrooms (Pleurotus ostreatus), with mean levels of 528 µg/g and 118 µg/g respectively (wet mass)25.

Placental transfer of SEN and EGT in a beluga fetus

We noted, for the first time, evidence of transplacental transport of both antioxidants by analysing the tissues of a fetus serendipitously found in a pregnant female. SEN and EGT were present in all fetal tissues but were not detected in fetal blood (Supplementary Table 3). Although SEN and EGT concentrations were notably lower and less variable than those in adults, with values well under 1 µg/g in all tissues, our results nevertheless indicate transplacental transfer of both compounds and their widespread distribution in the fetus.

Correlations between SEN and EGT

Positive and significant correlations were observed between SEN and EGT across all tissues except in brain and muscle (Fig. 3a–g). It should be noted that we had access to limited brain tissue samples (n = 5) compared to other tissues.

Fig. 3
Fig. 3The alternative text for this image may have been generated using AI.
Full size image

Correlations between concentrations (µg/g) of ergothioneine and selenoneine in skin (a), kidney (b), liver (c), brain (d), intestine (e), muscle (f), and whole blood (g) of adult belugas hunted in Quaqtaq, Nunavik, 2018–2019. R represents the Pearson correlation coefficient.

Associations between SEN and EGT in tissues had not been investigated prior to this study. Since (1) SEN and EGT share the same transporter (ETT) for cellular uptake5 (2) accumulate in all sampled tissues, (3) are strongly correlated in most tissues (Fig. 3a,b,c and g), and (4) can both be produced by specific microorganisms when selenium is present14,27, these observations suggest that ETT expression likely drives the distribution of these compounds among beluga tissues. ETT is highly conserved in all animals species and responsible for the absorption and distribution of SEN and EGT throughout the body2,5,28. Accumulation of these dietary antioxidants is generally highest in tissues submitted to high levels of oxidative stress and inflammation, such as liver, kidneys, erythrocytes, or eye lens29. The consistently greater concentration of EGT over SEN across tissues, including the skin, may indicate that ETT has a higher affinity for EGT than for SEN, that the beluga diet provides more EGT than SEN, or that EGT is eliminated more slowly than SEN. The large interindividual variability in SEN and EGT concentrations could be due to factors such as dietary habits, migration patterns or beluga population, as SEN and EGT are known to accumulate through the diet1,30.

SEN, EGT in skin layers and ETT expression

Concentrations of SEN and EGT across skin layers dissected from four belugas harvested in 2018–2019 are presented in Fig. 4a and b. A consistent increasing gradient of SEN and EGT concentrations from the inner to the outer layers was observed in all whales. The lowest mean concentrations were found in the dermis (7.54 µg/g SEN and 41.9 µg/g EGT), followed by intermediate concentrations in the stratum spinosum (17.2 µg/g SEN and 78.3 µg/g EGT for lower stratum spinosum/stratum basale and 23.2 µg/g SEN and 100 µg/g EGT in upper stratum spinosum) and the highest mean concentrations were reported in the stratum corneum (31.5 µg/g SEN and 161 µg/g EGT) (Fig. 4b).

Immunofluorescence labeling of ETT was performed on fresh skin samples collected from four additional whales harvested during the fall of 2023 (Fig. 4c and Supplementary Fig. 2). The strongest expression of the transporter was noted in rete ridges from the deeper layers of the epidermis (lower stratum spinosum and stratum basale), in fibroblasts from the papillary dermis as well as in the deeper layer of the dermis for all animals (Fig. 4c). ETT expression was absent in the upper epidermal layers (Fig. 4c).

ETT localisation in the skin suggests that the transporter is responsible for the increasing concentrations of SEN and EGT from the dermis to the epidermis. Stratum basale cells containing the antioxidants progressively migrate to the outermost part of the skin as they replenish the sloughing epidermal layer31. The increase in cellular density from the dermis to the epidermis is also important to consider, as the stratum corneum is the thinnest and densest layer of the skin32. Thus, higher concentrations observed in the epidermis may partly reflect differences in tissue composition and density, rather than solely ETT localisation or physiologically directed redistribution. Our results in the beluga are consistent with human data indicating that ETT is expressed in keratinocytes from the proliferative basal layer of the epidermis33.

As the stratum corneum of the skin, which exhibits the highest concentrations of SEN and EGT, can easily separate from the rest of the skin, some samples may have lost this layer prior to analysis. To mitigate this, triplicate skin sections were sampled from each whale, and mean values were used. Moreover, samples from pectoral or dorsal fins were excluded from the present analyses as these sites were previously shown to contain higher SEN concentrations11 due to a higher proportion of epidermis.

Fig. 4
Fig. 4The alternative text for this image may have been generated using AI.
Full size image

Concentrations of selenoneine and ergothioneine (in µg/g) across dissected skin layers of four beluga whales (a, b) 1: stratum corneum, 2: upper stratum spinosum, 3: lower stratum spinosum and stratum basale, 4: dermis (n = 4). Blubber was removed, as indicated by the scissors. Immunofluorescence detection of ETT in beluga skin (c). Images show nuclear DNA stained with DAPI (blue), ETT immunolabeling (red) and the superposition of both (merge) for each skin stratum from one whale. Scale bar = 50 μm.

SEN and EGT as skin protective agents

The presence of SEN and EGT in specific skin layers raises physiological questions as to why these external layers, which are consistently shed through molting, contain such high concentrations of these valuable antioxidants that are constantly lost to the environment. Molting is an important process for many whale species and might even explain the migration patterns of species that molt annually such as belugas34,35,36. Molting is an indicator of skin health in beluga whales as it contributes to a healthy microbiome and helps eliminate parasites and algae37,38. Molting is believed to protect against biofouling39 mercury accumulation32 and sun damage39.

Wagemann and Kozlowska32 previously reported that mercury concentrations in beluga skin increase outwardly in a pattern similar to that observed for SEN and EGT in the present work. They suggested that skin shedding and molting may serve as a mechanism for belugas to eliminate mercury. As the stratum corneum is lost in the molting process, this may function as a pathway to excrete mercury from the body32. This could be compared to a similar process in humans where mercury is known to be excreted in hair and nails40 or in bird feathers, such as the giant petrel, in the form of MeHg20. Furthermore, as there is increasing evidence that SEN is involved in sequestering and detoxifying MeHg in different tissues and species5the co-localisation of SEN and mercury in the layers suggests a unique interaction between the two compounds.

Martinez-Levasseur, et al.41 showed that whales are susceptible to sun damage and UV exposure, with factors such as skin color, time spent at the surface, and geographic latitude influencing the incidence of skin lesions and sun damage. Given that adult belugas are white and spend considerable time in shallow water or near the surface, especially in summer35,42, they are particularly vulnerable to UV photodamage, thus necessitating protective mechanisms. Since EGT has been shown to mitigate the effects of UV radiation and reactive oxygen species (ROS) in human skin, it is plausible that it may provide similar benefits to belugas33,43. Vitamin C, a well-known antioxidant that protects against UV photodamage and promotes skin health in humans44 has been identified in beluga skin45. Notably, vitamin C concentrations are significantly higher in the epidermis compared to the dermis45 mirroring the distribution patterns of SEN and EGT in the present study. Elevated concentrations of all three antioxidants may afford efficient protection against sun damage at these high latitudes. Interestingly, Inuit populations exhibit a lower prevalence of skin cancer compared to other populations46,47 despite significant exposure UV radiation through the practice of traditional activities on the land, including in winter when the snow albedo is very high. It is tempting to speculate that their traditional diet, which comprises beluga mattaaq and other SEN and EGT-rich country foods, resulting in exceptionally high blood levels of both antioxidants13 could be responsible for their apparent protection from skin cancer.

Conclusion

Our results corroborate the exceptionally high concentrations of SEN in beluga skin from our previous report21 and indicate even higher concentrations of its sulphur isologue EGT in this tissue. ETT-driven distribution of these antioxidants to keratynocytes in the regenerative epidermal layer suggests their possible involvement in protecting the skin against environmental stressors that may induce oxidative damage and lead to skin cancer. EGT and SEN could act as key antioxidants not only for whales, but also for coastal human populations who traditionally hunt these animals and may also benefit from a high intake of these dietary bioactive compounds.

Our findings are particularly significant for Inuit populations in Canada, where beluga whales play a central role in cultural practices and traditional diets. By shedding light on the nutritional and protective qualities of beluga, especially the skin, which contains high concentrations of vitamin C45 as well as both SEN and EGT, this research deepens our understanding of the health benefits associated with traditional Inuit foods, which are central to food security and intergenerational knowledge transmission11.

Methods

Sample collection

Samples were obtained from 14 beluga whales (Delphinapterus leucas) harvested by Inuit hunters during the subsistence hunt conducted along the coast of Quaqtaq (Nunavik, Canada) in Fall 2018, Spring 2019, and Fall 2019.

Total body length was recorded to the nearest 0.1 cm from nose to tail (Supplementary Table 1). Samples collected included liver (n = 13), kidneys (n = 13), brain (n = 5), muscles (n = 12) and intestines (n = 9). Skin samples from all individuals (n = 14) were collected from the dorsal area behind the blowhole, with three technical replicates per individual; replicate concentrations were averaged. Blood (n = 14) was collected post-mortem from the abdominal cavity into EDTA-treated tubes. One fetal specimen was collected whole, and the same tissues as those sampled in adults were dissected later in the laboratory. All samples were stored individually in plastic bags and readily frozen at − 20˚C. Additional details regarding the individual whales are provided in Supplementary Table 1.

Reagents and standards

All solutions were prepared using Milli-Q (ultrapure) water (18.2 M Ω cm, Millipore Bedford, MA).

LC-MS grade acetonitrile and methanol were obtained from Omnisolv (Omaha, NB). Hexane Optima grade was from ThermoFisher (Fair Lawn, NJ). Ammonium formate (99.995%; trace metals basis), dithiothreitol (DTT; ≥ 99%), and formic acid (≥ 95%; reagent grade) were purchased from Sigma-Aldrich (St. Louis, MO). L-ergothioneine, L-S-methyl-ergothioneine and L-ergothioneine-d9 were purchased from TRC (Toronto, ON, Canada). L-S-methyl-ergothioneine-d9 was produced as described in Achouba et al.16. L-selenoneine, L-Se-methyl-selenoneine, L-77Se-selenoneine and L-77Se-methyl-selenoneine were synthesised according to methods described in the international patent application PCT/CA2023/050197.

For immunofluorescence, water-based embedding medium (Tissue-Tek optimum cutting temperature (O.C.T.) compound) for frozen tissue specimens was purchased from Sakura Finetek Inc (Torrance, CA), formaldehyde solution 37%, from laboratory MAT (Quebec, QC) and fluoromount-G mounting medium from SouthernBiotech (Birmingham, AL). Bovine serum albumin, phosphate buffer saline (PBS) 1X and glycine (ultrapure) were obtained from Wisent Inc. (Saint-Jean-Baptiste, QC). Tween 20 (analytical reagent grade) and 4′,6-diamidino-2-phenylindole (DAPI) were from ThermoFisher (Fair Lawn, NJ). ETT/SLC22A4 mouse polyclonal antibody (A01) (H00006583-A01) was purchased from Cedarlane (Burlington, ON) and Alexa Fluor 647 conjugated goat anti-mouse IgG (H + L) polyclonal antibody from Jackson Immunoresearch Laboratories (West Grove, PA).

SEN, EGT, and methylated metabolite analysis by ID-LC-MS/MS

Blood samples were analysed according to the method previously developed and validated by Achouba, et al.16. Briefly, a 50 µL aliquot of blood was spiked with isotopically labeled internal standard solution containing 77Se-SEN (10 µg/mL), 77Se-methyl-SEN (1.0 µg/mL), EGT-d9 (50 µg/mL), and S-methyl-EGT-d9 (2.0 µg/mL) and mixed with 200 µL of a 50 mM DTT aqueous solution. The resulting mixture was then filtered through a 10 kDa cutoff 0.5 mL centrifugal filter (EMD Millipore, Omaha, NB) at 12 000 g for 45 min at room temperature (RT). A 20 µL aliquot of the filtrate was mixed with 180 µL of mobile phase A (described below) and the resulting solution was analysed using isotope dilution-liquid chromatography-tandem mass spectrometry (ID-LC-MS/MS).

Beluga tissue extraction and analysis were optimised based on the blood protocol. For skin samples, the adipose tissue (blubber) was removed prior to processing. The remaining tissue was delipidated using hexane, and the residual hexane was evaporated to dryness before homogenisation. For skin layer analysis, skin samples (n = 4) were cut horisontally into four visually distinct layers (Fig. 4b): the stratum corneum, thin cuticle easily removed on the surface of the skin; upper stratum spinosum, white layer absent of pegs; lower epidermal layer (lower stratum spinosum and stratum basale), darker layer with visible rete pegs; dermis, pink harder layer from which the blubber was previously removed. Blubber was not analysed in this study as previous analyses in our laboratory failed to reveal detectable concentrations of either SEN or EGT in this tissue. Internal tissue samples were used without any processing.

Approximatively 350 mg of tissue was weighed in a 2 mL reinforced polypropylene tube pre-filled with 2.8 mm ceramic beads (Omni International, Kennesaw, GA). To each sample, 200 µL of an isotopically labeled internal standard solution containing 77Se-SEN (7 µg/mL), 77Se-methyl-SEN (0.77 µg/mL), EGT-d9 (25 µg/mL), and S-methyl-EGT-d9 (1.0 µg/mL) was added, followed by 400 µL of a 50 mM DTT aqueous solution. Samples were then homogenised for four cycles of 30 s at 6 m/s using an Advanced Bead Mill Homogeniser (BeadRuptor Elite, Omni International, Kennesaw, GA). Homogenised samples were then centrifuged at 14 000 g for 10 min at 4 ˚C. The supernatant (extract) was transferred into a 10 kDa cutoff 0.5 mL centrifugal filter and centrifuged under the same conditions for 30 min. The extraction steps were repeated twice by adding 500 µL of the DTT solution to the pellet. A 50 µL aliquot of the combined filtrates was diluted in 200 µL of the DTT solution and vortex mixed. Finally, a 20 µL aliquot of the resulting solution was mixed with 180 µL of mobile phase A (described below) prior to analysis by ID-LC–MS/MS.

For both blood and tissue analysis, the chromatographic separation was performed using an Acquity I-Class UPLC system (Waters, Milford, MA), equipped with an Acquity HSS T3 column (150 × 2.1 mm, 1.8 μm), operated at 30 °C with a flow rate of 0.4 mL/min. A binary gradient of 0.1% formic acid with 10 mM ammonium formate (A) and acetonitrile (B) was used: 0–1.6 min (100% A), 1.6–2 min (linear to 10% A), 2–3.6 min (10% A), 3.6–8 min (100% A). The injection volume was 10 µL in full loop mode. The system was coupled to a Xevo TQS micro triple quadrupole mass spectrometer (Waters) using electrospray ionisation (ESI) in positive mode.

Validation of the analytical method

The sample preparation protocol was optimised to accommodate these new matrices, ensuring minimal matrix effects and achieving low limits of detection (LODs).

Representative chromatograms for each ion transition are presented in Supplementary Fig. 3. Chromatographic separation was consistent across all beluga matrices, with elution times of 1.40 min for EGT, 1.55 min for SEN, 1.91 min for S-methyl-EGT, and 2.03 min for Se-methyl-SEN.

LOD was established by repeatability (16 replicates) of beluga muscle tissue. Analytical performance data (Supplementary Table 4) indicate LODs of 0.5 µg/g for EGT and 0.04 µg/g for SEN, while LODs for the methylated metabolites were an order of magnitude lower, at 0.004 µg/g.

Due to the unavailability of commercially available certified reference materials as quality control (QC) for these compounds, inter-day precision (% CV) was assessed using pooled tissue samples: beluga skin for the high-concentration QC and seal muscle and liver for the low-concentration QC (n = 17 for each). For EGT, % CV values were below 10% at both concentration levels, showing high instrumental precision. SEN showed similarly low % CV in the high QC sample, whereas values in the low QC sample approached the LOD, resulting in a higher % CV (18%). A similar trend was observed for S-methyl-EGT and Se-methyl-SEN: both compounds were below the LOD in the low QC samples, and in the high QC samples, concentrations approached the LOD, leading to higher % CVs (28% and 15%, respectively) (Supplementary Table 4).

Recovery and matrix effects were evaluated for each analyte across all beluga tissues following the standardised procedure of Bienvenu et al.48. Global matrix effects, depending on the analyte, ranged from 98% to 106%, while instrumental matrix effects varied between 96% and 102%, indicating minimal impact on ionisation efficiency and quantitative accuracy of the analytes during mass spectrometry analysis. Recovery was above 70% for all matrices and effectively compensated by the internal standards employed for isotope-dilution quantification of analytes (Supplementary Table 4).

Immunofluorescence localisation of ETT in beluga skin

Fresh beluga skin samples were obtained from four additional animals during the Fall of 2023, thanks to the collaboration of hunters in Quaqtaq and biologists at the Department of Fisheries and Ocean Canada. Samples were placed on ice immediately after collection and transported by plane to Quebec City. Upon arrival to the laboratory, 1.5 cm width pieces of each skin sample were embedded in Tissue-Tek O.C.T Compound and kept at −80 °C until further use.

Rabbit cornea were used as a positive control to confirm the primary antibody specificity to ETT as it is expressed in abundance in the corneal epithelium49 (Supplementary Fig. 2a). Rabbit heads from young non-albino rabbits were obtained from a local slaughterhouse. Eyes were enucleated and dissected to isolate the cornea. OCT embedded corneas were kept at -80 °C until further use.

From the cryopreserved tissues, 5 μm thick cross-sections were obtained using a Leica MC3050 cryostat, and heat-fixed (30 min at 37 °C) on a microscope slide. Sections were then fixed in 3% formaldehyde for 15 min at RT, permeabilised with 0.1% Tween 20 in PBS for 15 min at RT, and incubated in blocking buffer (5% bovine serum albumin, 0.05% Tween 20, and 0.3 M glycine in PBS) for 1 h at RT. They were subsequently incubated with ETT/SLC22A4 mouse polyclonal antibody (A01) diluted 1:200, overnight at 4 °C. After washes with 0.1% Tween 20 solution, secondary antibody incubation was performed with Alexa Fluor 647 conjugated goat anti-mouse IgG (H + L) polyclonal antibody diluted 1:2500, for 1 h at RT. Nuclear DNA was counterstained with 0.33 µg/ml of DAPI in PBS for 15 min at RT. Slides were mounted with Fluoromount-G Mounting Medium and signals were visualised using a Zeiss Axioimager Z2 microscope coupled with a Zeiss Axiocam MRm Rev 3 Monochromatic Digital Camera. Specific binding of the primary antibody to ETT and its ability to detect ETT in beluga tissues were validated (Supplementary Fig. 2).

Statistical analysis

All analyses were performed using R v4.2.250. Concentrations below LOD values were replaced by LOD/2 prior to statistical analyses. Concentrations of analytes across different tissues were compared using a one-way analysis of variance (ANOVA) followed by the post-hoc Tukey pairwise comparison analysis based on the package “stats” (version 3.6.2) and a confidence interval of 95% (Supplementary Table 5a-d). Each ANOVA was performed following logarithmic transformations to satisfy the normality and homoscedasticity assumptions, which were tested using diagnostic plots and a Shapiro-Wilk test. Correlations were tested among analytes for each of tissue by the Pearson correlation coefficient using the “stats” package. All statistical tests were two-sided and were conducted without excluding any samples. All results are expressed as arithmetic means since the normality assumption was satisfied once concentrations were log transformed.