Evaluation of vitamin E requirement and food palatability in rabbits fed a purified diet with a high fish oil content

Transcription

1 64 Laboratory Animals (990) 4, 64-7 Evaluation of vitamin E requirement and food palatability in rabbits fed a purified diet with a high fish oil content P. M. VERSCHUREN, U. M. T. HOUTS MULLER & J. L. ZEVENBERGEN Unilever Research Laboratorium Vlaardingen, Olivier van Noortlaan 0, 333 AT Vlaardingen, The Netherlands Summary The vitamin E requirement of rabbits fed a semi-synthetic diet containing high amounts of fish oil was studied. Three groups of 5 rabbits were fed fish oil diets containing, respectively, 50, 00 and 500 mgkg vitamin E. Moreover diet palatability was evaluated by using different levels of grass meal: O 5, and %, respectively. Incorporation of % grass meal in the diet was sufficient to achieve acceptance of the fish oil diet. Increased vitamin E intake resulted in a dose-related rise in vitamin E levels in serum, blood platelets, liver and adipose tissue. The higher vitamin E intake was reflected by a twofold increase of vitamin E in serum, platelets and adipose tissue, and a tenfold increase in the liver. The adipose tissue revealed histopathological changes of yellow fat disease, mainly in the lowdose vitamin E group. In the liver microgranulomas of lipofuscin-laden macrophages were seen. Vitamin E was found to decrease but not to prevent the formation of these lesions. The results indicate that protection of marine oils against in vivo oxidation is problematic in the rabbit. It is questionable whether in this animal vitamin E is an adequate biological anti-oxidant for very long chain n-3 fatty acids. Keywords: Vitamin E requirement; Rabbits; Fish oil feeding Correspondence to: U.M.T. Houtsmuller, Unilever Research Laboratorium Vlaardingen, PO Box 4, 330 AC Vlaardingen, The Netherlands. Received 6 December 988; accepted 9 September 989 Fish oil is considered to play a preventive role in the development of cardiovascular disease, which is the leading cause of death in Western societies(herold & Kinsella, 986).Consequently, it is advisable to increase the dietary intake of fish oils. Fish oil contains high amounts of verylong-chain polyunsaturated f ltty acids (0: 5, n-3 and : 6, n-3), which are extremely susceptible to auto-oxidation and peroxidation in vivo. Therefore, a higher demand of biological antioxidant may be envisaged. In the present study we have investigated the ability of vitamin E to prevent in vivo auto-oxidation. Feeding semi-synthetic diets to rabbits is difficult.this holds particularly for diets containing fish oil, since rabbits have an aversion to the odour and taste of fish. In our experience food rejection is a common observation in such studies (unpublished results). In general, it was possible to habituate rabbits to a semi-synthetic diet by gradually replacing the standard conventional rabbit pellets by the pelleted experimental diet. The feeding of fish oil, however, frequently led to reduced food acceptance or rejection, and to histopathological changes in the liver. In a previous rabbit-feeding trial, in which the effect of fish oil among various oils on atherosclerosis was studied, it was observed that the livers of fish oil-fed animals had pronounced granulomas composed of lipofuscin-laden macrophages (Hornstra et 0., 983). As previously mentioned, fish oil is very prone to auto-oxidation. Addition of sufficient antioxidants is a prerequisite in order to avoid lipid peroxidation during diet preparation and animal feeding. Moreover, when feeding these fatty

2 Vitamin E and fish oil in rabbits 65 acids to animals, sufficient biologically active anti-oxidants are necessary to avoid lipid peroxidation in vivo. Vitamin E is one of the major lipid-soluble anti-oxidants present in biological systems. It is a membrane-associated vitamin and functions as a radical scavenger, effectively inhibiting the peroxidation of cellular membranes. The ability of vitamin E to protect membrane lipids from peroxidative damage is described by Tappel (965). The liver lesions we found in previous rabbit studies were possibly caused by suboptimal vitamin E levels in the diets. The literature only provides information on required anti-oxidant levels in conventional rabbit diets. Adams (987) prescribes a level of 40 mg tocopherol per kg pelleted conventional rabbit diet. However, the fat level in these diets normally ranges from 5 to 400 fat. It is well known that the anti-oxidant requirement of mammals depends on the total amount and type of fat. The present feeding trial with rabbits served two purposes. First, to define a method for rabbits to accept fish oil-containing semi-synthetic diets by varying the amount of grass meal and, secondly, to define an adequate level of vitamin E to avoid lipid peroxidation in vivo by incorporation of different vitamin E concentrations in diets containing fish oil. Materials and methods Animals Fifteen female Dutch belted rabbits of about 8 weeks old were obtained from Ranch Rabbits (Crawley Down, West Sussex, UK) and housed individually in stainless steel wire-bottomed cages (580 x 440 x 370 mm) under conventional conditions (temperature 8 ± ce, humidity 40-70% and artificial day-night cycle of h). Water was available ad libitum. Diets During a -week pre-treatment phase all animals were fed conventional rabbit pellets (LK-Ol rabbit pellets, Hope Farms, Woerden, The Netherlands). The pellets contained 3, 3% fat Table. Composition of the semi-synthetic rabbit diet (gio 000 kcal) Ingredient Group Group Group 3 Casein 6 0' Maize starch 05 '7 b 05' Fat mixture 43,0' 43,0 43,0 Sawdust 60 0 d 60,0 60,0 Grass meal I' 5' 3,0 6,0 Mineral mix 5'0 f Yitamin mix (}l! 0 0 dl-a-toxopherolacetate (mgkg) (50) (00) (500) '4, kcal.g-'; DMY, Yeghel, The Netherlands. b4 lkcal.g- l ; NY Honig's Artikelen, Koog aid Zaan, The Netherlands. '9,3 kcal.g-'; Fish oil, Unimills BY, Zwijndrecht, The Netherlands; Sunflowerseed oil, Union NY, Antwerp, Belgium; Cacao butter, Chempri BY, Raamsdonksveer, The Netherlands. dsawdust and grass meal both sterilized, Broekman Instituut, Someren, The Netherlands. 'Minerals mglooo kcal diet: MgHPO. 3H 0 630; KH PO. 95; KCI 5965; CaCO) 750; CH)COONa 300; MnSO. H 0 54; Fe-citrate (l9ojofe) 90; Cu-citrate (35OJoCu) '4; Zn-citrate (3OJoZn) 35; KIO) 0'3; CoSO f Yitamins mgloookca] diet: choline 95; myo-inositol 95; p-amino-benzoic acid 47; thiamin (monocitrate) ]8; pantothenic acid (Ca-salt) ]6'5; nicotinic acid 45; riboflavin 7; pyridoxine 7; folic acid 3; vitamin K3 '; biotin 0'5; vitamin A acetate (35IUmg) 4; cholecalciferol (vitamin D3, 80Umg) 5. and 76 mg vitamin E per kg diet. Subsequently the animals were allocated by weight to 3 groups of 5 animals each. The diets were gradually changed to pelleted semi-synthetic diets containing 40% of the digestible energy (en%) as a fat mixture of fish oil, mainly menhaden type (0 en%), cacao butter (7' 5 en%), and sunflowerseed oil (' 5 en%). In order to accustom the animals to these diets, the proportion of the semi-synthetic diet to the conventional diet was increased every weeks for 6 weeks. Each increase represented a,5% replacement of the conventional diet. Special attention was paid to the refining and storage of the fish oil, which resulted in an oil without the typical odour of fish. Directly after refining butyrylhydroxytoluene (BHT) was added to the oil as anti-oxidant at a concentration of 00 mgkg. Subsequently, the oil was stored at - 0 c e under nitrogen until use. The three treatment groups received, respectively, 50, 00 and 500 mg a-tocopherol acetatekg diet. In addition grass meal was

3 66 Verschuren, Houtsmuller, & Zevenbergen Table. Fatty acid composition of the experimental diet Fatty acid en% Fatty acids weight % 4: 0-7 SAFA" 4-9 6: MUFA b 8-3 6: I 8 PUFA' 9 9 8: : 8,9 8: 6 8: 3 0 0: 0 0'3 0: 0,3 0: : 0 0-\ : 0-4 : 6 ' n - 6 fatty acids 6 n - 3 fatty acids 5-3 saturated fatty acids 7 3 'saturated fatty acids; bmono-unsaturated fatty acids; 'polyunsaturated fatty acids. added to the three experimental diets in a concentration of I 5, 3 and 6 g000 kcal, respectively. The semi-synthetic diets were prepared weekly and stored at - 0 C until use. Freshly thawed diets were supplied ad libitum to the animals three times a week. The composition of the semisynthetic diets and the fatty acid composition of the experimental fat mixture are given in Tables I and. At the start of the feeding trial the peroxide value (PV) was determined of both the fat mixture and the complete diets, i.e. freshly prepared and after the maximal storage time (in practice 7 days at - 0 C followed by 3 days at room temperature). Observations All animals were checked daily for general health and weighed weekly. Daily food intake was measured every weeks at the end of each accustomization step over a period of 4 days. Two weeks after changing to the 5% and 50% diet replacement levels, blood samples were taken from the ear vein for vitamin E determinations. In week 8, which was weeks after reaching the 00% experimental diet level, blood was collected from the ear vein for haematology. Subsequently, all animals were killed under anaesthesia by exsanguination via aorta cannulation. Blood samples were collected from the aorta to determine vitamin E concentration in both serum and platelets. All carcasses then were subjected to a detailed necropsy. An assessment was made of the intra-abdominal fat deposition. Samples of liver and adipose tissue were taken for vitamin E analysis. Prior to fixation liver weights were recorded. Spleen, liver, mesenteric lymph nodes, and a sample of adipose tissue from the gonadal fat depot were fixed in 0% buffered formalin for subsequent histological examination. After processing of the formalin-fixed material according to routine procedures, paraplast sections were stained with Harris' Haematoxylinazophloxine and Masson's trichrome stain. Moreover, sections of liver and adipose tissue were stained with periodic acid Schiff (PAS) and with carbol fuchsin (Ziehl-Neelsen) for the detection of lipofuscin. Unstained sections were studied with a fluorescence microscope (Zeiss). All sections were read under code to avoid reader bias. Formalin-fixed samples of the liver were processed and examined in a Philips EM 400 T electron microscope at 60 UV. Vitamin E analysis Samples of serum and homogenized liver were briefly mixed with ethanol containing antioxidant to denaturate proteins. The lipophylic vitamins were extracted with two volumes of hexane containing a-tocopheryl acetate as internal standard (50AgmI). The hexane was removed under nitrogen and the residue was dissolved in about 5AIchloroformmethanol (l : ). Subsequently the samples were analysed by RP-HPLC on a Lichrospher RP 8(5), 50x 4, 6 mm column with a mixture of methanol! water tetrahydrofuran (97: 3 : 5) at l' 0 ml! min. The eluate was monitored at 87 nm and the results were calculated by comparing the relative peak areas using reported extinction coefficients (Catignani & Bieri, 983).

4 Vitamin E and fish oil in rabbits 67 Results Diet analysis The PV of the fat mixture was <. In the freshly prepared diets a PV of 4 ± O' 3 was determined, which increased only slightly up to 6 ± O' after the longest storage period, indicating good stability. Observations on living animals During the 8 weeks of feeding the rabbits appeared healthy. One animal participating in group 3 was an outlier in many aspects. It had a poor condition and showed only marginal food intakes during the weeks prior to necropsy and was therefore considered in iii health; all results for this animal were excluded from calculation of the mean values. Food intake decreased during the feeding period (Table 3). This was partly due to the Table 3. Food consumption (gday, average values ± SEM, n=s) Week Group J Group Group 3* I O±'6 04±4'8 08±4'4 3 04±5'4 99±6' 0±' ± ± 3,8 93± ±4 0 8 ±3'7 79±' 9 54±4'8 6±4'6 50± 6,5 II 55 ± 3,9 53±4'9 68± ± ± 5,8 50± ± 5,8 35 ±4'5 9±3' *n=4. gradual increase of the energy density of the diet during the adaptation procedure. Moreover, a number of animals showed short periods of food rejection, in particular directly after the interim blood collections, which coincided with the diet changes. Consequently, body weight showed erratic fluctuations towards the end of the habituation period (Fig. ). The average number of days per animal on which a diminished food intake was recorded during the I8-week feeding period was 38, 5 and 6 for the groups, and 3, respectively. The animals fed the 3 glooo kcal grass meal (corresponding to 0,0) showed acceptable food intakes. Haematology and clinical chemistry No relationship between the dietary treatments and blood parameters was found with regression analysis. All the vitamin E levels determined in the rabbits were correlated with the dietary intake of vitamin E (Table 4). The vitamin E values of Table 4. Vitamin E levels in serum (JLgml), blood platelets (LgI06), liver <l'gg) and adipose tissue (JLgg). (Average values ± SEM) Sample Group I Group Serum 4'0±0'58 Platelets 3' 3 ± 0 Liver 9 ± 3 6 Adipose tissue 46, 6 ± 8, 98 *n=4; tn=3. 6'5± \ \0 5'5±0'47 40'8±3'8 57,9 ± 5,94 Group 3* 8'6± \ 00 9'0± 0' ± 55 4 l'4± 5 7t bodyweoigh\ I kg.0.8 increoaseo' '! , Fig. l. Body weight of rabbits during the feeding period (average ± SEM; n = 5) group I;., group ; 0. group timelweeks '" '" '" '" '" ""' :.:..;.:.:-.... ~I(-_. ~ OL--_-'- -'-- --' ~_~ o vitamin E indieot I(mg.kg") Fig.. Increase in vitamin E levels in tissues, relalive to the group receiving the lowest dose. serum;., platelets; 0, liver; 0, adipose.

5 68 Verschuren. Houtsmuller. & Zevenbergen Table 5. Body weight and liver weight (average values ± SEM. n = 5) Parameter Group J Group Group 3 * Body weight (g) Liver weight (g) Relative liver weight (gioog bw) 7 ± 49'3 43,9 ± 3,94 '56±O'097 87±5'9 45,6 ± 4, 6 '50±O'05 776±55'9 34'6±O'94 95±O 076 *n=4. Table 6. Microscopic pathology: incidence of findings (n = 5). The absence of a numeral indicates that the finding specified was not identified Organfinding Liver fatty deposition slight moderate severe lipogranulomas + t (Zn positive) + (Zn positive) 3 + (Zn positive) 4 + (Zn positive) 5 + (Zn positive) slight focal periportal fibrosis mononuclear cell infiltration Mesenteric lymph nodes pigmented macrophages some moderate much erytrophagocytosis focal oedema Spleen Haemosiderosis slight moderate severe Some macrophages in white pulp Adipose tissue Steatitis stage I stage II Group J Group Group I I I diet, whereas the liver appears to accumulate the excess, possibly in a storage function. Post mortem observations The relative liver weight was negatively correlated with the amount of dietary vitamin E (Table 5). No macroscopic evidence for vitamin E deficiency was found (i.e. yellow fat disease). Two animals from the 50 mgkg vitamin E group displayed a yellow liver. On histological examination, treatment-related changes were found in liver and adipose tissue (Table 6). The majority of the rabbits showed a fatty deposition in the liver, possibly as a consequence of the high-fat diet. The fat deposition was mainly present as large droplets. Lipopigmentladen macrophages forming microgranulomas were found, mainly In the periportal areas but also throughout the whole liver acinus (Fig. 3). *n=4; tscore I +-5+ indicates incidental-numerous. the interim blood collections have not been included in the Table. The results show that whereas the concentration of vitamin E in serum, platelets and adipose tissue seem to follow compa:rable saturation curves, in the liver a linear correlation with the dietary dose is found (Fig. ). This might indicate that the former three tend to saturate rather rapidly above a dose of mgkg Fig. 3. Microgranuloma (a) in periportal area of liver, surrounded by fat-containing hepatocytes (b).

6 Vitamin E and fish oil in rabbits 69 microlipogranulomas were numerous in the lowdose vitamin E group (50 mgkg) and occurred only incidentally in the high-dose vitamin E group (500 mgkg). The group fed 00 mgkg vitamin E showed an intermediate position with respect to these changes. Lipofuscin accumulation was also seen in lymph node macrophages (paracortex and medulla) in all experimental animals. However, the degree of this lipofuscin accumulation was not related to the vitamin E concentration in the diet. Fig. 4. Acid-fastness of macrophages in microgranuloma the liver (Ziehl-Neelsen stain). of Fig. 6. Autofluorescence of lipofuscin in interstitial macrophages of adipose tissue..,.'ig. 5. Autofluorescence of lipofuscin in macrophages in lhe liver. r The identity of these inflammatory cel\s as macrophages was proved by electron microscopy. The pigment was acid fast, PAS-positive (Fig. 4) and showed a bright yellow autofluorescence (Fig. 5). These histochemical characteristics pointed to lipofuscin, a product of polymerization of peroxides of unsaturated fatty acids with proteins and other cellular material, possibly originating from cell debris (Pearse, 97). Slight lipofuscin accumulation was also found in individual hepatocytes throughout the liver. The Fig. 7. Degenerated fat cells (D) surrounded by inflammatory cells (Ziehl-Neelsen stain).

7 70 Verschuren, Houtsmuller, & Zevenbergen Fig. 8. Degenerated fat cell (D) with macrophages and Iymphoc)'tes (Ziehl-Neelsen stain). The adipose tissue lesions were found to be indicative of the development of yellow fat disease. The majority of animals in group and showed focal accumulation of interstitial lipofuscin-laden macrophages (autofluorescent and acid fast) in the adipose tissue (Fig. 6). This was considered to be an initial stage of yellow fat disease (steatitis stage I). In addition, next to the interstitial macrophage reaction, multifocal degeneration of fat cells was found in 5 rabbits. This phenomenon was characterized by degenerated fat cells which were surrounded by a corona of inflammatory cells (steatitis stage II) (Figs 7 and 8). The presence of the steatitis was related to the vitamin E level in the diet. Discussion Under certain conditions, feeding high-fish oil-containing semi-synthetic diets to rabbits is possible. First, it is necessary to use a carefully refined fish oil, having only a minimum of undesirable odour. Moreover, it is a prerequisite to stabilize the refined and deodorized oil properly by an adequate antioxidant, and to store the treated oij at low temperature under nitrogen in order to prevent lipid peroxidation. We applied special refining conditions followed by BHT-addition, which resulted in an almost odourless and tasteless product. Secondly, it is necessary to gradually replace the conventional rabbit diet by the semi-synthetic diet, while continuously monitoring the food intake. We found that by means of a step-bystep replacement of the conventional rabbit diet by the fish oil-containing semi-synthetic diet, the food rejection problems can be controlled. Thirdly, the diet intake is improved by incorporation of grass meal. Our findings indicate that although incidental food rejections are not fully prevented, our method of diet habituation results in satisfactory acceptance of the semi-synthetic fish oil diet. Fish oil-containing diets with insufficient amounts of antioxidants lead to rapid autoxidation of the oil and the occurrence of effects in the animal which are not due to the fish oil per se (Fritsche & Johnson, ] 988). However, not every antioxidant has anti-oxidant activity in vivo. So precautions must also be taken in order to avoid in vivo lipid peroxidation. Conventional rabbit pellets, usually low-fat, contain about 40 mg of a-tocopheryl acetate per kg diet (Adams, ]987). The present study reveals that at this vitamin E level, the experimental diet with of the digestible energy from fish oil causes obvious histopathological changes in liver and adipose tissue. Thf. presence of abundant microlipogranulomas in the liver and early changes of yellow fat disease in the gonadal fat pad are the consequences of in vivo autoxidation, leading to lipofuscin formation caused by insufficient protection vitamin E. The latter finding was also described in other species by Danse and Steenbergen- Botterweg (976) and Danse and Verschuren (978). The severity of the changes is negatively correlated with the amount of vitamin E in the diet. Nevertheless, the formation of lipogranulomas could not be fully prevented in our study. In a previous -year rabbit study on the effects of fish oil, sunflowerseed oil and linseed oil, the adverse effects were restricted to the fish oil-fed group only (to be published). All dietary groups involved in this study received ]00 mgkg vitamin E. The present results suggest that even a dietary

8 Vitamin E and fish oil in rabbits 7 dose as high as 500 mgkg vitamin E does not prevent lipid peroxidation of the marine oil, although the level in the liver followed the dietary concentration. This is in contrast to findings of Meydani et al. (987), who studied the effect of the type of dietary fat on the tocopherol status in mice. They found that increasing tocopherol levels in diets of mice fed fish oil (in contrast to corn oil and coconut oil) did not increase the liver tocopherol level. They concluded that this was due to either an interaction of fish oil and tocopherol at the gut level, or to an enhanced post-absorptive utilization of tocopherol. Our findings suggest that a significant interaction at the gut level is unlikely, but an enhanced postabsorptive utilization is possible. Since in our experiment the hepatocyte vitamin E levels are in line with the dietary vitamin E concentrations, it may be that lipid oxidation products are formed in a very early stage in the adaptation to the semi-synthetic diet. During this period the total dietary vitamin E level was still significantly lower. Another possibility is that the lipofuscin found in the liver was originally formed outside the liver and subsequently stored in this organ. This is supported by the early accumulation of lipofuscin in Kuppfer cells of the liver as observed in vitamin E deficient rats (Danse & Verschuren, 978). The adipose tissues of the animals fed high vitamin E levels showed no signs of yellow fat disease. On the basis of the results of this experiment, in particular the apparent saturation curves of serum, platelets and adipose tissue, it can be concluded that 50 mgkg vitamin E in a semisynthetic diet with a high fish oil content is too low to prevent excessive lipid peroxidation. The vitamin E concentration should be within the range of 00to 500 mgkg purified diet, preferably around 00 mgkg. Since obviously rabbits fed fish-oil-containing diets have a high demand for vitamin E, in contrast to those fed diets containing vegetable oils (Homstra et al., 983), it seems appropriate to incorporate a high vitamin E concentration from the very start of the feeding period, i.e. from the very first day of the habituation phase, even when the amount of fish oil fatty acids is still very low. Further studies on anti-oxidant activity are required to elucidate the question whether in the rabbit vitamin E is the most suitable biological antioxidant for very-long-chain fish oil fatty acids. Acknowledgments Thanks are due to GAA Kivits, WGL van Nielen and P Moret for their input in the vitamin analyses. FH Wijnen, W Tuitel, W van Oort and lec van Wijk are thanked for support in the animal care. The skilful technical assistance of WG Timmer, E Haddeman and AD van der Kooij is gratefully acknowledged. We thank Dr LHJC Danse, (Dutch) National Institute of Public Health and Environmental Protection, for his scientific advice. References Adams CE (987)The laboratory rabbit. In The UFAW Handbook on the Care and Management of Laboratory Animals 6th edition. Harlow: Longman Scientific & Technical Catignani GL & Bieri JG (983) Simultaneous determination of retinol and a-tocopherol in serum or plasma by liquid chromatography. Clinical Chemistry 9,708-7 Danse LHJC & Steenbergen-Botterweg WA (976) Early changes of yellow fat disease in mink fed a vitamin E deficient diet supplemented with fresh or oxidized fish oil. Zentralblatt fur Veterinaer Medizin 3, Danse LHJC & Verschuren PM (978)Fish oil induced yellow fat disease in rats. Veterinary Pathology 5, 4-4 Fritsche KL & Johnson PV (988) Rapid autoxidation of fish oil in diets without added antioxidants. Journal of Nutrition 8, Herold PM & Kinsella JE (986) Fish oil consumption and decreased risk of cardiovascular disease: a comparison of findings from animal and human feeding trials. The American Journal of Clinical Nutrition 43, Homstra G, Haddeman E, KJoezeJ & Verschuren PM (983) Advances in Prostaglandin, Thromboxane, and Leukotriene Research, volume (eds B. Samuelsson, R. Paoletti, & P. Ramwell). New York: Raven Press Meydani SN, Shapiro AC, Meydani M, Macauley JB & Blumberg JB (987) Effect of age and dietary fat (fish, Cornand coconut oils) on tocopherol status of C57BL6Nia mice. Lipids, Pearse AGE (97) Histochemistry, Theoretical and Applied 3rd edition, volume II. London: Churchill Tappel AL (965) Free radicals lipid peroxidation damage and its inhibition by vitamin E and selenium. Federation Proceedings 4, 73