This work was reported in a preliminary from as the proceedings of 11th International Symposium on the Cells of the Hepatic Sinusoid and their Relation to Other Cells, Tucson, Arizona, USA, 25-29 August, 2002 in Comparative Hepatology 2004, 3 (Suppl I) (http://www.comparative-hepatology.com/content/3/S1/S18)
Metabolites of vitamin A, termed retinoids, are a group of molecules that exert multiple and profound effects on a wide variety of biological processes such as cell proliferation, differentiation, morphogenesis, and tumorigenesis (Blomhoff, 1994; Blomhoff and Blomhoff, 2006). With the exception of vision, most or all of these effects seem to be mediated by retinoic acid isomers binding to one or several of six specific nuclear receptors (Chambon, 1996; Chawla et al., 2001; Balmer and Blomhoff, 2002; Mezaki et al., 2007, 2009). The observations that all vertebrate cell types express several of these receptors, and vitamin A-deficiency and -excess leads to abnormalities in a number of processes, including cellular growth and differentiation, reproduction and immune defense, suggest that a precise fine-tuning of retinoic acid synthesis and catabolism is essential for the function of most cells.
The liver is the main storage site of vitamin A in humans and animals (Ralli et al., 1941; Underwood et al., 1970; Raica et al., 1972; Schmitz et al., 1991). In physiological conditions, hepatic stellate cells (HSCs; also called as vitamin A-storing cells, lipocytes, interstitial cells, fat-storing cells, or Ito cells) store 50–80% of the total vitamin A in the whole body as retinyl palmitate in lipid droplets in the cytoplasm. HSCs also regulate both transport and storage of vitamin A (Wake, 1980; Blomhoff et al., 1985, 1991, 1992; Blomhoff and Wake, 1991; Blomhoff, 1994; Blomhoff and Blomhoff, 2006; Friedman, 2008; Blaner et al., 2009; Senoo et al., 2010, 2011).
More than 95% of vitamin A in the HSC is present in the form of retinyl esters packed together in cytoplasmic lipid droplets. The normal reserve of retinyl esters in HSCs represents an adequate supply of vitamin A for most individuals for several weeks or months (Blomhoff et al., 1990). This extensive storage of retinyl esters in HSCs, and the cells' ability to control mobilization of retinol, ensures a steady blood plasma retinol concentration of about 1–2 μM in spite of normal fluctuation in daily intake of vitamin A.
Some human cases of hypervitaminosis A have been reported among explorers and travelers in the arctic (Conway, 1906; Doutt, 1940; Rodahl and Moore, 1943; Rodahl, 1950; Biesalski, 1989) after eating polar bear liver. Typical symptoms of hypervitaminosis A are changes in central nervous system, with increased pressure of cerebrospinal fluid such as headache, dizziness, vomiting, loss of appetite, as well as effects on skin and mucous membranes such as dryness, peeling, cheilitis, brittle finger nails, loss of hair, and changes to bone such as swelling and periosteal apposition accompanied by pain in the bones (Biesalski, 1989; Myhre et al., 2003).
Vitamin A accumulates at near toxic concentrations in some arctic predators (Rodahl and Davies, 1943; Rodahl and Moore, 1943; Rodahl, 1949). Polar bears contain a high concentration of vitamin A in their liver, but the precise cellular localization remains to be identified. We performed this study to demonstrate cellular localization of vitamin A systematically in arctic animals, including top predators such as polar bear, arctic fox, bearded seal, and glaucous gull.
MATERIALS AND METHODS
The protocols for animal use described in this article herein were previously approved by the Animal Research Committee of each author's institute namely, Akita University Graduate School of Medicine, Japanese Red Cross Medical Center, and Faculty of Medicine of University of Oslo. All subsequent animal use adhered to the “Guidelines for Animal Experimentation” of each institute. After getting permission from the district governor of Svalbard, liver specimens were collected from three polar bears (Ursus maritimus), eight arctic foxes (Alopex lagopus), 14 bearded seals (Erignathus barbatus), six ringed seals (Phoca hispida), 13 glaucous gulls (Larus hyperboreus), five fulmars (Fulmarus glacialis), one Brünnich's guillemot (Uria lomvia), five puffins (Fratercula arctica), seven Svalbard reindeers (Rangifer tarandus platyrhynchus), and five Svalbard ptarmigan (Lagopus mutus hyperboreus). The polar bears came from Ny Ålesund and Hornsund in Svalbard (February and August 1998). The other animals came from the Svalbard archipelago near Longyearbyen (78° N, 15° E; August 1996–April 1998). We analyzed liver of 13 brown bears (Ursus arctos) from Jämtland, Gävleborg and Dalarna, Sweden, four red foxes (Vulpes vulpes) from Västergötaland, Sweden, and eight gray gulls (Larus argentatus) from Skåne, Sweden. We also collected livers from four polar bears in 2001, four polar bears, one hooded seal, and two glaucous gulls in 2006, and four polar bears in 2008, all from Scorebysund (Ittoqqortoormiit), Greenland.
High-Performance Liquid Chromatography
Each piece of organ was extracted with appropriate amounts of organic solvents as described previously (Tsuiki et al, 2007). The extract was subjected to high-performance liquid chromatography (HPLC). Retinol and retinyl esters were measured by the method described previously (Tsuiki et al, 2007). The retinoids were identified on the basis of retention time. The spectra of the peaks were monitored with a diode-array detector (L-2450; Hitachi, Ltd., Tokyo, Japan). The ratio of absorbances at 330 and 350 nm were compared with those of retinoid standards (Irie et al., 2004). The extraction and analyses of retinoids were performed under dim red light during experiments.
Gold Chloride Staining
Two blocks (3 × 3 × 2 cm3) from each organ were subjected to a specific vitamin A-staining technique, Wake's modification (Wake, 1971; Wake et al., 1986) of the gold chloride-staining method (Kupffer, 1876). Briefly, the blocks were immersed in 0.05% chromic acid for about 2 hr at room temperature. Sections were frozen and cut into 50-μm thick strips by a freezing microtome (Leitz, Germany) and placed in 0.05% chromic acid for 10 min. The sections were incubated in the gold chloride solution (1 mL of 1% gold chloride, 1 mL of 1% HCl, 98 mL of distilled water) at 18–20°C in total darkness. When the sections turned red-purple, they were dehydrated with ethanol and mounted on balsam (Senoo and Wake, 1988; Higashi and Senoo, 2003; Senoo et al., 2010, 2011).
Fluorescence Microscopy for Detection of Autofluorescence of Vitamin A
Two blocks (3 × 3 × 0.5 cm3) were immersed in 3.7% formaldehyde for 24 hr at 4°C in total darkness, and 20-μm thick sections were made with a freezing microtome (Leitz, Germany). The sections were examined with a Zeiss Axioskop 20FL (excitation filter BP365/12, barrier filter BP495/40) for the detection of rapidly fading green autofluorescence characteristic of vitamin A (Popper and Greenberg, 1941; Wake, 1980; Wake et al., 1986; Higashi and Senoo, 2003; Senoo et al., 2010).
Liver sections were stained with Sudan III, hematoxylin and eosin, and Ishii-Ishii's silver impregnation (Ishii and Ishii, 1965) methods as described previously (Wake, 1980; Wold et al., 2004; Senoo et al., 2010). For hematoxylin and eosin staining, several blocks (3 × 3 × 0.5 cm3) were immersed in 3.7% formaldehyde, dehydrated in a series of graded ethanol, and embedded in paraffin. For Sudan III staining and Ishii–Ishii's silver impregnation, several blocks were immersed in 3.7% formaldehyde and cut at 10-μm thickness with a freezing microtome.
Transmission Electron Microscopy
A block (3 × 3 × 2 cm3) from each organ was perfused with 1.5% glutaraldehyde in 0.062 M cacodylate buffer, pH 7.4, containing 1% sucrose for 1 or 2 min by injection through blood vessels which appeared on the cut surface of the block. After perfusion, tissue blocks (2 × 2 × 2 mm3) were postfixed in 2% osmium tetroxide in 0.1 M phosphate buffer, pH 7.4, for 2 hr at 4°C, dehydrated in a graded ethanol series, and embedded in Epon 812. Ultrathin sections were made with an ultramicrotome 2088-V (LKB) and stained with 2% uranyl acetate and 0.1% lead citrate. Thin sections were examined under a transmission electron microscope (JEOL-1200EX) at an acceleration voltage of 80 kV (Nagy et al., 1997; Imai and Senoo 1998; Imai et al., 2000). Thick sections were examined under a light microscope after staining with 1% toluidine blue containing 1% borax.
Morphometric Analysis using Transmission Electron Microscopy
HSCs were examined by counting the number of cells carrying lipid droplets in the cytoplasm with nucleus in perisinusoidal space in an ultrathin section (Weibel et al., 1966; Loud, 1968; Nagy et al., 1997). From each animal, 10 liver sections were randomly made and analyzed. Nine micrographs (magnification of 1000×) were studied per section (i.e., 3,900 mm2/micrograph). For determining the distribution of the lipid droplets in HSCs, the printed photomicrographs were scanned (Epson GT-6500 ART2) and lipid droplet areas were recognized with Adobe Photoshop 3.0 J software. The area of the lipid droplets was measured with NIH Image 1.58, and then statistically calculated with Microsoft Excel. Data are expressed as mean ± SE. Statistically significant difference was determined by the t-test (P < 0.05).
Hepatic Storage of Total Retinol in Arctic Animals
The amounts of vitamin A stored in livers of arctic animals are shown in Table 1. Median values are presented due to large individual variations. The liver of top predators from Svalbard, including polar bear, arctic fox, bearded seal, and glaucous gull, contained 18.3, 18.6, 4.7, and 6.8 μmol total retinol per gram liver, respectively. Polar bears from Greenland contained 33.5 μmol total retinol per gram liver (Table 1).
Table 1. Total retinol contents in livers of arctic animals
Concentration (μmol/g wet weight)
Total amount in organ (μmol)
Percentage of retinyl esters
Caught in Svalbard.
Caught in Greenland.
The liver weight for polar bears was estimated by assuming that the fraction of liver to body weight was similar to that of brown bears. Total retinol content (i.e., the sum of retinol and retinyl esters) was measured as described in “Materials and Methods.”
Other organs such as kidney, spleen, lungs, intestine, and muscle stored a small amount of total retinol (data not shown).
Vitamin A Staining and Autofluorescence from HSCs
As shown in Fig. 1, strong autofluorescence (Fig. 1A,C,E,G) and gold chloride staining (Fig. 1B,D,F,H) were present in HSCs of polar bear (Fig. 1A,B), arctic fox (Fig. 1C,D), glaucous gull (Fig. 1E,F), and bearded seal (Fig. 1G,H). Livers of ringed seal (Fig. 1I,J), Svalbard reindeer (Fig. 1K,L), and Brünnich's guillemot (Fig. 1M,N) also showed specific vitamin A autofluorescence and staining in HSCs. The intensity of vitamin A autofluorescence and staining in HSCs of ringed seal (Fig. 1I,J), Svalbard reindeer (Fig. 1K,L), and Brünnich's guillemot (Fig. 1M,N) was weaker than that of the top predators (Fig. 1A–H), and comparable or somewhat more intense than the staining in human and rat liver (data not shown). Black-stained gold chloride reaction was seen only on the surface of the lipid droplets in HSCs of polar bears (Fig. 1B) because lipid droplets (as shown in Fig. 4C) in the cytoplasm of HSCs that emanated autofluorescence of vitamin A were too large for the gold chloride solution to penetrate into the core of the droplet (Fig. 1A,B).
No specific staining or autofluorescence of vitamin A was seen in any of the other cell types in liver, including parenchymal cells (hepatocytes), liver sinusoidal endothelial cells (LSECs), and Kupffer cells (KCs; Fig. 1). Other organs, such as kidneys, spleen, lungs, intestine, and muscle, emanated almost no vitamin A autofluorescence or staining (data not shown).
Distribution of Lipid and Extracellular Matrix Components
To demonstrate the distribution of lipid in the vitamin A droplet in the liver, sections from polar bear liver were stained by Sudan III (Fig. 2). This technique identified regularly distributed lipid droplets in HSCs of polar bear liver. The distribution of lipid droplets was similar to the distribution of vitamin A autofluorescence emanated from HSCs (Fig. 1A) and vitamin A-stained droplets (gold chloride technique) in HSCs (Fig. 1B) of polar bear liver. Very few lipid droplets were observed in other cell types, namely, hepatocytes, LSECs, and KCs.
Vitamin A lipid droplets in the cytoplasm of HSCs of polar bear (arrows in Fig. 3A), arctic fox (arrows in Fig. 3D), and glaucous gull (arrows in Fig. 3G) were recognized in the hematoxylin and eosin-stained paraffin sections. One or two large lipid droplets in the cytoplasm were readily observed in HSCs of polar bear liver (arrows in Fig. 3A).
Extracellular components, such as collagen, proteoglycan, glycosaminoglycan, and glycoprotein, distributed in the sinusoidal wall of the liver were demonstrated by the silver impregnation method of Ishii and Ishii (1965; Fig. 3B,E,H). At lower magnification (Fig. 3C,F,I), neither hepatic fibrosis nor liver cirrhosis was demonstrated.
Nuclear deviation of hepatocytes on sinusoidal surfaces was obvious (Fig. 3A,B,D,E,G,H), as reported previously (Sato et al., 2001).
Electron Microscopic Demonstration of Hepatic Ultrastructure
We looked the hepatic ultrastructure in arctic animals by means of transmission electron microscopy. In livers of arctic fox (Fig. 4A,B), polar bear (Fig. 4C), glaucous gull (Fig. 4D), bearded seal (Fig. 4E), and ringed seal (Fig. 4F), HSCs in the perisinusoidal space of Disse stored lipid droplets in their cytoplasm. One or two large vitamin A lipid droplets were demonstrated in the cytoplasm in HSCs of polar bear (Fig. 4C), bearded seal (Fig. 4E), and ringed seal (Fig. 4F); several medium-sized lipid droplets in arctic fox (Fig. 4A), and many small-sized ones in glaucous gull (Fig. 4D) were seen. The lipid droplets occupied almost all the cytoplasm and only few cell organelles such as rough endoplasmic reticulum and mitochondria were observed. Fixatives penetrated into some lipid droplets (gray color; L in Fig. 4A,C), but did not penetrate well into other lipid droplets (white color, L in Fig. 4C). Lipid droplets often indented the nucleus (Fig. 4A,C,E,F). These morphological characteristics are in general consistent with that of HSCs with high amounts of retinyl esters (Wake, 1971, 1988; Yamada and Hirosawa, 1976; Blomhoff and Wake, 1991).
Morphometric Analysis of Cell Number and Lipid Droplets in the Liver of Arctic Mammals and Birds
As demonstrated by morphometric analysis of electron micrographs, there were no significant differences among the cell numbers of HSCs per hepatocytes of the animals examined (Table 2). The lipid droplet areas were, however, significantly larger in the liver of polar bear, arctic fox, and bearded seal compared to that of normal rat livers. The lipid droplet areas of ringed seal and glaucous gull were similar to that of rat livers but lipid droplet area in Svalbard ptarmigan liver was significantly smaller than that of rat liver. These data suggest a relationship between the area of the lipid droplets and the amount of stored retinyl ester determined with HPLC. The data also indicate that storage capacity of retinyl ester is mainly due to the capacity of each HSC, not to the number of HSCs.
Table 2. Morphometric analysis of cell numbers and lipid droplets
Stellate cells in % of parenchymal cells
Area fraction of lipid droplets in stellate cells (μm2)
Mean ± SE
Mean ± SE
The morphometric quantitation was performed as described in “Materials and Methods.” The results are presented as means ± SE; n = 3 in each group.
Total retinol values in arctic top predators are much higher than all other arctic animals studied as well as their genetically related continental top predators. The values are also high compared to normal human values and experimental animals like mouse and rat (200–600 nmol per gram) (Raica et al., 1972; Nagy et al., 1997; Rosales et al., 1999). Values are also much higher than livers of food-producing animals like calf, ox, lamb, pig, and chicken, which contain about 300–600 nmol per gram (Howell and Livesey, 1998).
In mammals, most of the vitamin A is normally located in HSCs that constitute less than 1% of the liver's cell number and mass. This is a specialized cell type that stores retinol as long-chain fatty acid esters in cytoplasmic lipid droplets (Wake, 1971; Blomhoff and Wake, 1991). The normal role of HSCs is to ensure ample access of the vitamin during periods with low dietary intake. HSCs also play an important role as a detoxifying mechanism by reducing the amount of free retinol, which is toxic for cells, following intake of large doses of vitamin A (Wake, 1971; Blomhoff and Wake, 1991; Blomhoff, 1994; Myhre et al., 2003).
As we demonstrated in this study, the localization of stored vitamin A in the arctic animals was essentially the same as that published previously in normal rat and human liver, namely HSCs (Wake, 1971; Wake, 1988; Blomhoff and Wake, 1991; Senoo et al., 2010, 2011). In hypervitaminosis A in human and rat (Stimson, 1961; DiBenedetto, 1967; Rubin et al., 1970; Muenter et al., 1971; Hruban et al., 1974; Russell et al., 1974; Jacques et al., 1979), vitamin A spills out from HSCs into the portal area (Glisson's sheath) (Wake, 1971; Wake, 1988; Blomhoff and Wake, 1991, plasma, and other organs (Rodahl and Davies, 1943; Rodahl and Moore, 1943; Rodahl, 1949, 1950; Biesalski, 1989). In polar bear, arctic fox, glaucous gull, and bearded seal, the liver has vitamin A levels that are comparable to hypervitaminosis A in human and rat, the portal area did not show vitamin A autofluorescence or positive gold chloride staining (Fig. 1). Other organs contained low levels of total retinol (data not shown). Plasma concentration of total retinol in the arctic animals was also essentially the same as that of humans and animals in marginal nutritional status (data not shown). These observations suggest that these arctic animals have developed a higher vitamin A storage capacity in HSCs than most other animals.
Polar bear livers were previously examined by morphological and biochemical methods (Leighton et al., 1988). Large lipid droplets were in cells located in perisinusoidal areas by oil red-O, and hematoxylin and eosin staining methods, and suggesting that these cells were HSCs (also called Ito cells). The autofluorescence of vitamin A was not detected and gold chloride reaction was unstable in the sections that were used for that study because the specimens were kept for 1.5 to 2.0 months before preparation. In the present study, autofluorescence of vitamin A and positive gold chloride reaction are clearly and stably shown because we used fresh specimens. Thus, we investigated vitamin A storage localization systematically not only in liver in polar bears but also in other arctic animals, including top predators.
Meanings of Storage of Vitamin A in Arctic Top Predators
As mentioned above, our findings demonstrate that the arctic animals have a very high vitamin A-storing capacity in HSCs. Such a high capacity for vitamin A storage may have been crucial for the survival and evolutionary adaptation of the arctic predators to the extremely high levels of vitamin A that accumulate through the food web in such animals. HSCs are central in the pathogenesis of liver fibrosis (Senoo and Wake, 1985), which is a characteristic sign for vitamin A-, and xenobiotic-toxicity. No such pathological signs were detected so far in any of the animals studied.
Every animal is closely linked with a number of other animals living around it, and these relations in an animal community are largely food relations (Elton, 1927). These interrelations between animals are not difficult to study if the following four principles are realized. The first is that of food chains, the second principle is the size of food, the third principle is that of niches (the animal's place in its community, its relations to food and enemies, and some extent to other factors), and the fourth idea is that of pyramid of numbers in a community. By the fourth idea is meant greater abundance of animals at the base of food-chains, and the comparative scarcity of animals (predators) at the end of such chains. Elton visited Spitsbergen in 1921 and reported food chains among the animals (Summerhayes and Elton, 1923; Elton, 1927). Elton emphasized significance of the conservation of variety (biodiversity) (Elton, 1958).
Here, from our present study, we propose as a hypothesis that HSCs play pivotal roles in the development, establishment, and maintenance of food web and food chain in the arctic. Without huge capacity of storage of vitamin A in HSCs in the arctic top predators, these animals may suffer from hypervitaminosis A after intake of the large amount of vitamin A through the food chain, and cannot stay in the top of the food web there.
The authors are grateful to Jørn Eldar Fortun, Anders Friberg, Trond Øfstaas, Hjelmer Hammeken, Jørgen Arke, and Georg Bangjord for samples of fulmars (Fulmarus glacialis) and also of some of the bearded seals (Erignathus barbatus) and a ringed seal (Phoca hispida), and Arne Söderberg at Svenska Jägareförbundet for samples from brown bears (Ursus arctos). We also thank the University Courses at Svalbard Longyearbyen, Svalbard (especially Dr. Rolf Langvatn) for their generous assistance by allowing us to use their laboratory facilities in Longyearbyen, the Norwegian Polar Institute, and the District Governor at Svalbard.
They also thank Dr. Kenjiro Wake (Emeritus professor of Tokyo Medical and Dental University; Liver research Unit, Minophagen Pharmaceutical Company) for his valuable discussions, and Naosuke Kojima and Kiwamu Yoshikawa (Department of Cell Biology and Morphology, Akita University Graduate School of Medicine), Kari Holte and Heidi L. Wold (Department of Nutrition, Institute of Basic Medical Sciences, Faculty of Medicine, University of Oslo) for their excellent technical skills.