In chickens, as with all animals, body fatty acids are derived from dietary uptake, de novo synthesis and/or bioconversion. Among the different fatty acid classes, n-3 long-chain PUFA (n-3 LCPUFA) are of particular interest due to their beneficial role on human health(Reference Ruxton, Calder and Reed1). In vertebrates, α-linolenic acid (18 : 3n-3) and linoleic acid (18 : 2n-6) cannot be biosynthesised de novo, but are derived from the diet and subsequently can be bioconverted to longer and more unsaturated n-3 and n-6 LCPUFA, respectively. This in vivo bioconversion of n-3 LCPUFA includes endoplasmic Δ-6 desaturation, chain elongation and Δ-5 desaturation of the precursor 18 : 3n-3 to EPA (20 : 5n-3), which is subsequently converted to docosapentaenoic acid (22 : 5n-3) by chain elongation. The final metabolite, DHA (22 : 6n-3), is synthesised by chain elongation, Δ-6 desaturation and peroxisomal β-oxidation of 22 : 5n-3(Reference Leonard, Pereira and Sprecher2). The n-6 pathway involves the same enzymes and conversion steps as the n-3 pathway, with arachidonic acid (20 : 4n-6) being the major metabolite of dietary 18 : 2n-6.
The absolute amount of 18 : 3n-3 intake is of prime importance to the efficiency of conversion to the LCPUFA(Reference Griffin3). However, the production of n-3 LCPUFA, particularly 22 : 6n-3, from 18 : 3n-3 is limited in the human body(Reference Burdge and Calder4, Reference Brenna, Salem and Sinclair5). Isotopic tracer studies showed that the conversion of 18 : 3n-3 to 22 : 6n-3 is about 1 % in infants and substantially lower in adults with, however, important differences between men and women(Reference Burdge and Calder4, Reference Brenna, Salem and Sinclair5). Therefore, it is often postulated that the n-3 LCPUFA are semi-essential and should be sufficiently provided by the diet. In order to optimise the fatty acid composition of foods derived from farmed animals, knowledge of their LCPUFA metabolism is required. To study PUFA metabolism, several in vivo and ex vivo methods have been applied in different species. Isolation of cells or tissue microsomes and incubation with labelled fatty acids is a common ex vivo method(Reference Brown6). Besides the need for sophisticated analyses and expensive reagents, other drawbacks of the ex vivo methods are that some controlling factors are missing and that this approach is restricted to one tissue(Reference Turchini, Francis and De Silva7). Therefore, to gain an accurate insight into the whole body (WB) metabolism of PUFA, a proper in vivo method is desired. Different in vivo techniques appropriate to WB fatty acid metabolism have been detailed elsewhere(Reference Cunnane8). Recently, a WB fatty acid balance method has been suggested as a reliable method to study the overall PUFA metabolism in fish(Reference Turchini, Francis and De Silva7). The WB fatty acid balance method comprises a feeding experiment, quantification of the initial and final WB fatty acid composition and the excreted fatty acids. Thereafter, the partitioning of fatty acids among excretion, accumulation, conversion to longer chain and more unsaturated metabolites and mitochondrial β-oxidation is computed. To the best of our knowledge, no research has examined the different fates of n-3 and n-6 PUFA metabolism (accumulation, conversion to longer chain and more unsaturated metabolites and mitochondrial β-oxidation) in poultry simultaneously by a WB approach. The rationale for the present study was to quantify the fate of individual dietary fatty acids in broiler chickens as a function of the dietary fatty acid source, and the animal factors age and sex through a WB fatty acid balance method.
Materials and methods
Animals and diets
The experiment was carried out according to the guidelines of the Ethics Committee of Ghent University (Belgium) for the humane care and use of animals in research. A total of 400 1 d-old chickens of the Ross308 strain were obtained from a local hatchery (Belgabroed, Merksplas, Belgium). The birds were randomly allotted to eight floor pens with fifty males or females per pen, with each pair of pens randomly selected to receive either one of the four experimental diets. The experiment was accomplished in two periods of age (days 7–21 and days 21–42). On day 7 and day 21, the birds were individually weighed, and six birds per pen around the average live weight were selected and placed in metabolic cages with two birds of the same sex per cage and three replicate cages per diet × sex combination. Birds in the cages received the same diets as in the pens. Evolution of live weight gain was comparable in the cages and in the pens. Feed and water were provided ad libitum. Feed intake and body weight were recorded for the caged birds during the respective period of study. Four diets based on wheat and soyabean meal and differed only in the fat source were fed (Table 1): palm fat (Palm, 8 % palm fat, SFA source); soyabean oil (Soya, 5 % palm fat+3 % soyabean oil, 18 : 2n-6 source); linseed oil (Lin, 5 % palm fat+3 % Lin, 18 : 3n-3 source); Fish oil (Fish, 5 % palm fat+3 % fish oil, n-3 LCPUFA source). The fatty acid profile of the diets is given in Table 2. Differences between diets for average live weight at the start and end of the two periods were minimal, except for a somewhat lower average live weight on day 21 and day 42 for the Fish diet.
Composition of the basal experimental diet
CP, crude protein.
* Exogenous phytase from Ronozyme P-5000.
† Bio Feed Wheat enzymes; NSP enzyme preparation from Ronozyme WX.
‡ The vitamin/mineral premix provided the following quantities (mg/kg of diet): retinyl acetate 4; cholecalciferol 0·07; dl-α-tocopheryl acetate 50; menadione 2·5; thiamin 2·2; choline chloride 650; riboflavin 7·5; pantothenic acid 38; pyridoxine 5·5, cyanocobalamin 0·035; nicotinic acid 13; biotin 0·20; folic acid 1; ethoxyquin 35; butylated hydroxytoluene 25; iodine 2; Co 1; Se 0·4; Cu 25; Mn 60; Zn 70; Fe 45.
Fatty acid profile of the diets
Palm, palm fat diet; Soya, soyabean oil diet; Lin, linseed oil diet; Fish, fish oil diet; –, not detected; ΣSFA, sum of SFA; ΣMUFA, sum of MUFA.
Sample collection and sample preparation
On days 21 and 42, after overnight fasting, caged birds were killed by cervical dislocation. Since the present study was part of a larger study, data from the analysis of separate anatomical compartments were used and pooled. Blood was collected, and the carcasses were dissected in seven body compartments: (1) complete deboned skinless thigh muscle; (2) skinless breast muscle; (3) liver; (4) heart; (5) brain; (6) abdominal fat pad (fat surrounding cloaca and adjacent muscles); (7) blood and carcass trimmings, i.e. skin, wings, feet, bones, neck, offal, feathers, visceral organs and remainders, named the rest compartment. The parts composing the rest compartment were minced together in a 20 litre three-bladed cutter for 10 min (Moled HK20/L, REX-MASCIINEN, Pinneberg, Germany). Samples per compartment were pooled for the two birds per cage, weighed, vacuum packed and stored at − 20 °C until analysis. From days 7 to 21 and days 21 to 42, faeces were totally collected from the cages, weighed, homogenised, and a representative sample from each cage was freeze-dried and stored at − 20 °C until analysis. Information on the fatty acid composition of different body compartments can be found in a recent report by Poureslami et al. (Reference Poureslami, Raes and Huyghebaert9).
Fatty acid analyses
Lipids were extracted from feed, body compartments and excreta using the method of Folch et al. (Reference Folch, Lees and Sloane Stanley10) upon mincing and homogenisation. Samples were then transmethylated, and fatty acid methyl esters were analysed by GC (HP6890, Brussels, Belgium) on a CPSil88 column for fatty acid methyl esters (100 m × 0·25 mm × 0·2 μm, Chrompack, Middelburg, The Netherlands) according to the method described by Raes et al. (Reference Raes, De Smet and Demeyer11). Fatty acids were identified by comparing their retention times with those of the corresponding standards (Sigma, Brussels, Belgium) Nonadecanoic acid was used as internal standard to obtain quantitative data of the fatty acid contents (mg/100 g tissue). Before WB fatty acid balance calculations, a theoretical correction was made for fatty acid methyl esters based on the response correction factors discussed by Ackman(Reference Ackman12).
Whole body fatty acid balance calculations
The different fates (excretion, body accumulation, elongation, desaturation and mitochondrial β-oxidation) of dietary 18 : 2n-6 and 18 : 3n-3 were calculated according to the method described in detail by Turchini et al. (Reference Turchini, Francis and De Silva7, Reference Turchini, Francis and De Silva13) and Turchini & Francis(Reference Turchini and Francis14). Summarising the methodology, the WB fatty acid balance was computed in distinct steps. Initially, the fatty acids in the diets, faeces and the initial and final carcasses (sum of fatty acid mass in seven body compartments) were quantified in mg. Knowing the fatty acid mass ingested and excreted, apparent digestibility of individual fatty acids was computed, and the net fatty acid intake (fatty acid absorption) was calculated. The difference between fatty acid accumulation in the WB and the net intake was considered as the total appearance or disappearance of a specific fatty acid. Next, the balance of n-3 and n-6 PUFA was computed. The quantity of fatty acid was first converted from mg to μmol of appeared/disappeared fatty acid per chicken WB. Then a backward calculation was made along each of the fatty acid metabolic pathways. The number of μmol of a longer chain or more unsaturated fatty acid that appeared was subtracted from the number of μmol of the previous fatty acid in the pathway. This allowed quantifying the bioconversion (i.e. elongation and/or desaturation) of an individual fatty acid expressed as μmol of fatty acid/g body weight per day. Fatty acid β-oxidation was subsequently computed after all computations along the pathway. In four cases (out of forty-eight), a small negative β-oxidation value was obtained (i.e. an appearance of 18 : 2n-6 or 18 : 3n-3). These values were turned from a negative value to zero before further calculation. It was verified that this did not affect the significance of the treatment effects. The equations and details for calculating the fate of individual fatty acids (% ex novo production, % body accumulation, % bioconversion, % β-oxidation) were previously described(Reference Turchini and Francis14).
Data analyses
The data were analysed using S-PLUS for Windows (version 6.1; Insightful, Seattle, WA, USA). A linear model was used to analyse the fixed effects of diet, age and sex and their interaction terms. In case of a significant diet effect, the mean values were compared with the Tukey's post hoc test (P < 0·05).
Results
In Table 3, the n-3 and n-6 PUFA content of WB broiler chicken is given. Feeding the Soya diet resulted in higher contents of 18 : 2n-6, 20 : 3n-6 and 20 : 4n-6 compared with the three other diets (P < 0·001), whereas the Lin diet induced the highest 18 : 3n-6 and 18 : 3n-3 contents compared with the other diets (P < 0·001). The 18 : 3n-3 content in Lin fed birds was 11-fold greater than the average of the other dietary treatments (P < 0·001). Birds fed the Fish diet recorded a substantially higher n-3 LCPUFA content (20 : 5n-3, 22 : 5n-3 and 22 : 6n-3) compared with the other diets (P < 0·001). The 20 : 5n-3, 22 : 5n-3 and 22 : 6n-3 content in Fish fed birds was 14·3-, 8·4- and 5-fold greater compared with the average of the non-Fish fed birds. There was a trend for an effect of diet on the total body fatty acid content (P = 0·051), with lower values for the Palm and Fish diets compared with the Soya and Lin diets. As expected, age and sex had a significant effect on the total body fatty acid content.
n-3 and n-6 PUFA content of whole body broiler chickens
Palm, palm fat diet; Soya, soyabean oil diet; Lin, linseed oil diet; Fish, fish oil diet; RMSE, root mean squares error; D, diet; A, age; S, sex.
a,b,c,d Mean values within a row with unlike superscript letters were significantly different (P < 0·05).
* Significant interaction terms at P < 0·05.
† Sum of SFA, MUFA and PUFA.
PUFA intake, apparent digestibility, total appearance/disappearance and fate of 18 : 2n-6 and 18 : 3n-3 are presented in Table 4. Apparent digestibility for 18 : 2n-6 was significantly higher for Soya fed birds compared with the other diets (P < 0·001). Apparent digestibility of 18 : 3n-3 was higher on the Soya and Lin diets than on the Palm and Fish diets (P < 0·001).
n-3 and n-6 PUFA intake, apparent digestibility, appearance/disappearance, accumulation, bioconversion and β-oxidation in broiler chickens
Palm, palm fat diet; Soya, soyabean oil diet; Lin, linseed oil diet; Fish, fish oil diet; RMSE, root mean squares error; D, diet; A, age; S, sex.
a,b,c,d Mean values within a row for diets with unlike superscript letters were significantly different (P < 0·05).
* Significant interaction terms at P < 0·05.
Disappearance of PUFA may imply elongation and desaturation to longer chain metabolites or utilisation of their carbon skeleton through β-oxidation for energy production. Disappearance of 18 : 2n-6 and appearance of 20 : 4n-6 were higher in Soya fed birds in contrast with the other dietary treatments (P < 0·001). Disappearance of 18 : 3n-3 and appearance of 18 : 4n-3 and 20 : 5n-3 were higher in Lin fed birds compared with the other diets (P < 0·001). Appearance of 22 : 5n-3 was similar for the Lin and Fish fed birds and higher than for the Palm and Soya diets (P < 0·001). Appearance of 22 : 6n-3 was higher on the Lin diet than on the Soya and Palm diets (P < 0·05 only for difference Lin v. Palm diet). Birds of the Fish group recorded a net disappearance of 18 : 4n-3, 20 : 5n-3 and 22 : 6n-3 (P < 0·001).
Birds fed the Lin diet had a higher 18 : 2n-6 accumulation (% of net intake) compared with the other diets (P < 0·01). Bioconversion (% of net intake) of 18 : 2n-6 to longer chain/more unsaturated fatty acids decreased in the order Palm>Lin>Soya>Fish (P < 0·001). The 18 : 2n-6 β-oxidation (% net intake) on the Lin diet was 2-fold lower than on the other diets (P < 0·001). Accumulation of 18 : 3n-3 was not significantly different among the diets. Birds fed the Fish diet did not demonstrate bioconversion of 18 : 3n-3 to 18 : 4n-3. However, 18 : 3n-3 bioconversion on the Palm and Soya diets was similar and 1·5-fold higher than on the Lin diet (P < 0·001). The 18 : 3n-3 β-oxidation on the Fish diet was 1·4-fold higher than on the other dietary treatments (P < 0·001).
The effect of age was significant for PUFA intake, apparent digestibility, appearance/disappearance (except of 22 : 5n-3) and the proportions of the net intake of 18 : 2n-6 and 18 : 3n-3 accumulated and β-oxidised. Birds slaughtered at day 42 of age, when compared with the 7–14 d age period, had higher values for PUFA intake, PUFA apparent digestibility and 18 : 2n-6 and 18 : 3n-3 accumulation, but lower values for PUFA appearance/disappearance and β-oxidation. The effect of sex on the above-mentioned variables was significant for 18 : 2n-6 and 18 : 3n-3 intake, 18 : 3n-3 bioconversion and 18 : 3n-3 β-oxidation. Males had a greater 18 : 2n-6 and 18 : 3n-3 intake and 18 : 3n-3 β-oxidation (P < 0·05), whereas females showed a higher bioconversion of 18 : 3n-3 to 18 : 4n-3 (P < 0·05).
Accretion of Δ-5 and Δ-6 desaturated fatty acids and accretion of elongated fatty acids are presented in Table 5. Conversion of 20 : 3n-6 to 20 : 4n-6 as an indication of apparent Δ-5 desaturation activity and conversion of 18 : 2n-6 to 18 : 3n-6 as an indication of apparent Δ-6 desaturation activity in the n-6 pathway were higher in Soya fed birds compared with the other dietary treatments (P < 0·001). The Δ-5 desaturation of 20 : 4n-3 to 20 : 5n-3 was higher in Lin fed birds compared with the other diets (P < 0·001). Similarly, Δ-6 desaturation of 18 : 3n-3 to 18 : 4n-3 and 22 : 5n-3 to 22 : 6n-3 was greater on the Lin diet compared with the other diets (P < 0·001). Elongation of 18 : 3n-6 towards 20 : 3n-6 was higher in Soya fed birds than in the other diets (P < 0·001), and elongation of 18 : 4n-3, 20 : 5n-3 and 22 : 5n-3 was higher in birds fed the Lin diet than in the other diets (P < 0·001). Apparent desaturation activity was higher at 7–21 d of age than at 21–42 d of age (P < 0·001). Apparent elongation activity on 18 : 3n-6, 18 : 4n-3 and 22 : 5n-3 was higher in birds slaughtered on day 21 compared with day 42 (P < 0·001), whereas elongation of 20 : 5n-3 to 22 : 5n-3 was not affected by age (P>0·05). Sex had no influence on the above-mentioned variables (P>0·05).
Accretion of Δ-5 and Δ-6 desaturated and elongated fatty acids in broilers
Palm, palm fat diet; Soya, soyabean oil diet; Lin, linseed oil diet; Fish, fish oil diet; RMSE, root mean squares error; D, diet; A, age; S, sex.
a,b,c,d Mean values within a row for diets with unlike superscript letters were significantly different (P < 0·05).
* Significant interaction terms at P < 0·05.
The apparent Δ-6 desaturase activity acting on 18 : 3n-3 and 18 : 2n-6 was plotted against their respective net intake both expressed as μmol/g per d (Figs 1 and 2). The apparent Δ-6 desaturase activity on 18 : 3n-3 showed a strong positive linear relationship with the 18 : 3n-3 net intake (R 2 0·98), indicating that the higher the dietary 18 : 3n-3 supply, the higher the resultant product (18 : 4n-3). The response of apparent Δ-6 desaturase activity to 18 : 2n-6 net intake was positive but curvilinear (R 2 0·66), showing that by enhancing substrate (18 : 2n-6) availability, the efficiency of Δ-6 desaturase (18 : 3n-6 production) increased only up to a certain level. The Δ-5 desaturase activity on 20 : 4n-3 manifested a strong positive relationship (R 2 0·99) with the summed 18 : 4n-3 and 18 : 3n-3 dietary supply (Fig. 3).
Apparent Δ-6 desaturase activity on 18 : 3n-3 in relation to the 18 : 3n-3 net intake (μmol/g per d) across four diets, age and sex. Linear regression equation: Y = 0·0030+0·0919X; R 2 0·98; root mean squares error = 0·019 (n 16). ●, Fish oil diet; 
, linseed oil diet; ■, soyabean oil diet; Δ, palm fat diet.
Apparent Δ-6 desaturase activity on 18 : 2n-6 in relation to the 18 : 2n-6 net intake (μmol/g per d) across four diets, age and sex. Broken-line quadratic regression equation: Y = 0·3167 if X>6·669 and Y = 0·3167 − 0·0107 × (6·669 − X)2 if X ≤ 6·669; R 2 0·66; root mean squares error = 0·058 (n 16). ●, Fish oil diet; , linseed oil diet; ■, soyabean oil diet; Δ, palm fat diet.
Apparent Δ-5 desaturase activity on 20 : 4n-3 in relation to the 18 : 4n-3+18 : 3n-3 net intake (μmol/g per d) across three diets, age and sex. Linear regression equation: Y = 0·0197+0·0728X; R 2 0·99; root mean squares error = 0·011 (n 12). , linseed oil diet; ■, soyabean oil diet; Δ, palm fat diet.
Significant interaction terms are mentioned in the tables, but these effects are not further discussed here since they were negligible compared with the main effects and did not compromise the main conclusions. Most of the significant interaction effects were diet × age effects. P-values for the interaction effects were considerably smaller compared with the main effects. In addition, the significant interaction effects were the result of scale differences in most cases, i.e. the difference between the two age groups differed among diets in magnitude but not in sign (and vice versa) in case of a significant diet × age effect. In few cases, the effect of age was opposite or absent in one diet versus the other diets, mostly when very low values occurred for a diet. The only marked diet × age interaction effect was for the digestibility of 18 : 2n-6 and 18 : 3n-3, which was approximately 10 % lower in the young compared with the older birds on the Palm diet, whereas the age effect was negligible on the other diets.
Discussion
The effect of diet and age on the fatty acid apparent digestibility in the present study is in agreement with previous studies in broiler chickens(Reference Krogdahl15–Reference Smink, Gerrits and Hovenier17). Many of the physiological functions necessary for lipid digestion are immature at hatch and develop over the next weeks. Bile salt secretion appears to be the first limiting factor for lipid digestion through the first few weeks post hatch(Reference Krogdahl15). In the present experiment, the apparent digestibility values of 18 : 2n-6, 18 : 3n-3 and 20 : 5n-3 are in the same range as recently reported specifically for broiler chickens(Reference Smink, Gerrits and Hovenier17). An earlier study described higher animal fat digestibility in female compared with male chickens(Reference Krogdahl15). A slightly higher PUFA digestibility in female birds fed the Fish diet was observed in the present study; however, this difference was not statistically different. It should also be kept in mind that the digestibilities in the present experiment may be overestimated due to PUFA microbial biohydrogenation in the hindgut(Reference Smink, Gerrits and Hovenier17).
The ingested PUFA face one of the three fates in the body, i.e. accumulation, conversion to longer chain/more unsaturated metabolites or β-oxidation. The WB fatty acid balance method allows quantifying the fate of the WB 18 : 2n-6 and 18 : 3n-3 relative to the net intake. Irrespective of the dietary treatment, broiler chickens seem to have a preference for accumulation of 18 : 2n-6 above 18 : 3n-3. Reviewing the literature, quantitative data on fatty acid metabolism in broiler chicken are scarce. Accumulation of 18 : 2n-6 and 18 : 3n-3 in WB level of growing pig fed a maize–soyabean meal basal diet (without added fat) was approximately 67 and 48 % of digestible fatty acid, respectively(Reference Kloareg, Bellego and Mourot18). In a WB fatty acid study with Murray cod fish fed with Lin, 40 and 53 % accumulation of net intake for 18 : 2n-6 and 18 : 3n-3 were respectively reported(Reference Turchini, Francis and De Silva19). These values are lower than for the 18 : 2n-6 and 18 : 3n-3 % net intake accumulation in Lin fed broilers in the present study. In a recent study with rainbow trout fed with fish oil, 91 and 60 % net intake of 18 : 2n-6 and 18 : 3n-3 were respectively deposited in the WB(Reference Turchini and Francis14). However, in Lin fed trout, the corresponding values were 70 and 58 % of net intake for 18 : 2n-6 and 18 : 3n-3, respectively. Therefore, it is possible to speculate that comparing chicken with rainbow trout both fed with fish oil, 18 : 2n-6 and 18 : 3n-3 accumulation is greater in fish than in chicken. Consequently, 18 : 2n-6 and 18 : 3n-3 accumulation appears greater in chicken compared with fish when fed Lin.
Across the four dietary treatments, the conversion rate of 18 : 3n-3 to 18 : 4n-3 was 1·8-fold higher than the conversion rate of 18 : 2n-6 to 18 : 3n-6 indicating a higher affinity of Δ-6 desaturase for 18 : 3n-3 than for 18 : 2n-6. A relatively lower 18 : 3n-3 bioconversion and increased β-oxidation rate in the Lin diet in contrast with Palm and Soya fed birds are in agreement with a previous report by Vermunt et al. (Reference Vermunt, Mensink and Simonis20), stating that diets rich in 18 : 3n-3 increase 18 : 3n-3 β-oxidation, coincident with decreasing its conversion rate to longer and more unsaturated n-3 LCPUFA. In a WB fatty acid study with Murray cod fish fed Lin diet, 4 and 8 % of 18 : 2n-6 and 18 : 3n-3 net intake were respectively converted to longer chain metabolites(Reference Turchini, Francis and De Silva19). Hence, it seems there is a larger 18 : 3n-3 conversion in Lin fed Murray cod fish than in Lin fed chicken in our experiment. Similar to our findings, rainbow trout fed with fish oil did not demonstrate conversion of 18 : 3n-3 to 18 : 4n-3 in WB(Reference Turchini and Francis14). Feeding vegetable oil rather than fish oil increases desaturase and elongase activity as well as upregulates liver fatty acid elongase and FAD5 gene expression(Reference Turchini and Francis14, Reference Miller, Bridle and Nichols21). This explains the higher conversion of 18 : 3n-3 to 18 : 4n-3 in the Palm, Soya and Lin diets (5- to 9-fold) compared with the Fish diet in the present study. The 18 : 2n-6 bioconversion in rainbow trout fed with fish oil(Reference Turchini and Francis14) was 1·8-fold higher than in chicken fed fish oil in the present study. Comparing rainbow trout fish fed with Lin with the Lin treatment in our experiment indicated that the 18 : 2n-6 conversion was 1·2-fold higher in chicken albeit the 18 : 3n-3 conversion was 2-fold higher in fish.
The higher 18 : 3n-3 bioconversion in female birds of the present study is in agreement with previous findings in other species, e.g. human and rat(Reference Marra and de Alaniz22, Reference Burdge and Wootton23). Both the role of steroid hormones and other sex differences in metabolism may be involved. The n-3 LCPUFA content in plasma and tissues of rats appear to be positively related to circulating concentrations of oestradiol and progesterone and negatively associated with circulating concentrations of testosterone(Reference Childs, Romeu-Nadal and Burdge24), and testosterone administration to rats strongly depressed Δ-5 and Δ-6 desaturase activity(Reference Marra and de Alaniz22). In addition, fat β-oxidation was found to be greater in men than in women in several studies(Reference Burdge and Wootton23, Reference Jones, Murphy and Stolinski25), corresponding to our finding that β-oxidation of 18 : 2n-6 and 18 : 3n-3 was higher in males than in females. Therefore, the availability of 18 : 3n-3 for conversion to metabolites may be higher in females than in males. However, caution is needed when comparing the present results with those from studies involving mature animals or human subjects, since the male chickens in the present study were still immature. It is thus not clear whether the observed differences are mainly due to an effect of sex hormones or due to differences in energy metabolism.
The WB fatty acid balance method applied in the present study provided an estimation of apparent in vivo Δ-5 and Δ-6 desaturase activity. Regression analysis demonstrated a linear response in Δ-6 desaturase activity in the n-3 pathway to an increasing supply of the substrate 18 : 3n-3. The same trend was observed for Δ-5 desaturase activity. On the other hand, the response in Δ-6 desaturase activity in the n-6 pathway to an increasing supply of 18 : 2n-6 was curvilinear. However, it should be noted that the range of net intake of 18 : 2n-6 was larger than for 18 : 3n-3, and that the curvilinear response could be partly due to the considerably lower Δ-6 desaturase activity on the Fish diet compared with a similar 18 : 2n-6 net intake values on the other diets.
Crespo & Esteve-Garcia(Reference Crespo and Esteve-Garcia16) reported that the total β-oxidation of n-6 PUFA was 29 and 11 % of net intake, and 29 and 14 % for n-3 PUFA, in linseed and sunflower oil fed broiler chickens, respectively. However, the individual fates of 18 : 2n-6 and 18 : 3n-3 were not reported. In growing pigs fed a maize–soyabean meal diet, 30 % of the digested Σn-6 fatty acid and 52 % of the digested 18 : 3n-3 were β-oxidised at WB level(Reference Kloareg, Bellego and Mourot18). The 18 : 3n-3 β-oxidation reported in the present study was 31 % (across diets, age and sex), and therefore it seems to be lower in broiler chickens in contrast with swine. In rat, Leyton et al. (Reference Leyton, Drury and Crawford26) measured exhalation of 14CO2 and demonstrated that the β-oxidation rate of 18 : 2n-6 and 18 : 3n-3 was 48 and 64 %, respectively. In human, over 30 % of labelled 18 : 3n-3 administered were used as energy source in man; however, lower values (22 %) were reported in women(Reference Burdge and Wootton23, Reference Burdge, Jones and Wootton27). Consistently, numerically lower β-oxidation values for 18 : 2n-6 and 18 : 3n-3 were also recorded in female chickens in the present study.
In fish, the β-oxidation of 18 : 2n-6 and 18 : 3n-3 was 29 and 8 % of net intake, respectively, at WB level of Murray cod fish fed with Lin(Reference Turchini, Francis and De Silva19); while in rainbow trout fed with fish oil, 18 : 2n-6 and 18 : 3n-3 β-oxidation was reported as 7 and 40 % of net intake(Reference Turchini and Francis14). Comparing our findings with the latter study, which was implemented with similar methodologies, reveals that 18 : 2n-6 β-oxidation in fish oil fed chicken is about 3- to 4-fold higher than in rainbow trout, while 18 : 3n-3 β-oxidation is approximately similar in the two species (40 v. 41 % of net intake). In Lin fed rainbow trout, the β-oxidation of 18 : 2n-6 and 18 : 3n-3 was 26 and 29 % of the net intake(Reference Turchini and Francis14). This suggests that 18 : 2n-6 β-oxidation in linseed fed chicken is 3-fold lower than in linseed fed rainbow trout. On the other hand, 18 : 3n-3 β-oxidation in linseed fed chicken from the present study corresponds with linseed fed rainbow trout fish (29 % net intake). It is at this point important to highlight some fundamental differences between fish and birds basic nutrition: fish are cold-blooded animals and have minor metabolic energy requirements (i.e. no energetic costs for maintaining the position and minor energetic costs for the detoxification of N containing compounds). Birds derive the majority of their energy requirements from dietary carbohydrates; while in fish, carbohydrates are very poorly utilised. Accordingly, the lipid content of diets for the two groups of animals is remarkably different (approximately 8 % in broiler chickens and approximately 25 % in trout). Despite these important differences, our comparisons depicted earlier are relative to the fate of specific fatty acids as percentage of net intake, and therefore these differences have to be considered levelled. Thus, these comparisons are sound and useful in providing a better understanding of PUFA metabolism in different farmed animals.
The stage of the growth and development influences fatty acid β-oxidation(Reference Kloareg, Noblet and Milgen28). In our experiment, β-oxidation of 18 : 2n-6, 18 : 3n-3, 18 : 4n-3, 20 : 5n-3 and 22 : 6n-3 was higher in 21 d-old birds compared with 42 d-old birds. This corresponds with the fact that young fast growing chickens have a higher metabolism rate compared with the same birds approaching slaughter age.
The WB fatty acid balance method provides information about the overall fatty acid metabolism in the body. The method is comparatively inexpensive, easy and feasible in animal research laboratories(Reference Turchini, Francis and De Silva7). However, there are some assumptions and limitations in this approach that should be considered. The method does not allow to study LCPUFA retroconversion. Likewise, fatty acid β-oxidation is estimated based on the fatty acid disappearance, while the use of fatty acids for the synthesis of non-fatty acid metabolites, e.g. hormones or PG, is not considered (though quantitatively extremely limited). However, it is important to underline that other methods employing labelled fatty acid techniques are accomplished with similar assumptions. The main difference between the WB method and other methods used for fatty acid metabolism studies (e.g. perfused isolated liver isotopical study) is the time frame(Reference Turchini and Francis14, Reference Francis, Peters and Turchini29). In fact, the WB method needs a longer time frame compared with the other methods. Consequently, the average enzyme activity (enzyme velocity relative to total enzyme product over time) is estimated in a longer time frame. Therefore, estimated apparent enzyme activity may deviate from the instantaneous enzyme activity measured at a given time, but it likely provides useful information relative to whole metabolism in growing animals, resulting in a reliable estimation of the apparent in vivo metabolic activity.
In consideration of the fact that n-3 LCPUFA (namely 20 : 5n-3, 22 : 5n-3 and 22 : 6n-3) are reportedly beneficial to human health(Reference Ruxton, Calder and Reed1, Reference Griffin3, Reference Brenna, Salem and Sinclair5), there is great interest towards finding possible ways to increase the n-3 LCPUFA content of farmed animals. In the present study, we recorded that in broiler chickens fed the Lin diet for 42 d, hence receiving the highest amount of 18 : 3n-3, a total of 5·8 % of the net intake of 18 : 3n-3 was eventually converted to n-3 LCPUFA. On a molar basis, the conversion of 18 : 3n-3 to 20 : 5n-3 and 22 : 6n-3 was 2·7 and 3·3 % on the Palm diet, 1·6 and 2·7 % on the Soya diet and 2·0 and 1·2 % on the Lin diet, respectively. In rainbow trout fed a Lin-based diet, it was reported that 8·8 % of the net 18 : 3n-3 dietary intake was bioconverted to n-3 LCPUFA(Reference Turchini and Francis14). Thus, in comparison to the values reported in human subjects(Reference Burdge and Calder4, Reference Brenna, Salem and Sinclair5), the 18 : 3n-3 to n-3 LCPUFA bioconversion capacity of broiler chicken and rainbow trout appears to be remarkably greater, particularly the synthesis of 22 : 6n-3. Two important conclusions can be obtained from this observation. Generalising, it can be speculated that the efficiency of the bioconversion of 18 : 3n-3 to n-3 LCPUFA seems to be inversely related to evolution (rainbow trout>broiler chicken>human). Secondly, the inclusion of Lin (or other sources of 18 : 3n-3) in animal feed results in beneficial net production of n-3 LCPUFA. Considering breast and thigh meat (without skin) as the only edible parts of the chicken carcass, and using their proportion of the body weight and their fatty acid composition(Reference Poureslami, Raes and Huyghebaert9), it was calculated that 100 g of meat (breast and thigh) provided 62 mg n-3 LCPUFA (20 : 5n-3, 22 : 5n-3 and 22 : 6n-3) on the Lin diet. On the other hand, the actual total intake of 18 : 3n-3 to produce the live biomass corresponding to 100 g of edible meat was 7581 mg. On the Fish diet, 100 g meat provided 198 mg n-3 LCPUFA, corresponding to an intake by the broiler chickens directly of 4284 mg n-3 LCPUFA. Although the supply of n-3 LCPUFA was 3-fold higher in meat from Fish fed chicken compared with Lin fed chicken, the latter meat did not require consumption of n-3 LCPUFA by the broiler chicken. Thus, from a human nutrition viewpoint, the present data suggest that it is more beneficial to consume chicken fed Lin than the practice of direct consumption of Lin or other sources of 18 : 3n-3 alone by human subjects, in consideration that the conversion efficiency of 18 : 3n-3 to n-3 LCPUFA in human subjects is lower than 0·80 %(Reference Wang, Harris and Chung30). Additionally, from an environmental viewpoint, the use of fish oil in chicken nutrition seems to be a relatively wasteful practice of this precious and limited natural resource.
In conclusion, the metabolism (elongation, desaturation and β-oxidation) of PUFA in broiler chickens is mainly affected by the dietary fatty acid source rather than the animal factors. Dietary provision of 18 : 3n-3 promotes the apparent elongation and desaturation activity on fatty acids in the n-3 pathway. Hence, 18 : 3n-3 bioconversion relative to the net intake of this fatty acid seems to be limited by the dietary supply. Feeding fish oil suppresses apparent elongase and desaturase activity, but it was responsible for significantly higher total deposition of n-3 LCPUFA. Apparent elongase and desaturase activity appeared to be higher in young broiler chickens (21 d) compared with those at conventional slaughter age (42 d). Sex has only a marginal effect on the PUFA metabolism in broiler chicken.
Acknowledgements
The authors wish to acknowledge the Iranian Ministry of Science and Technology for providing the first author the opportunity to do this research. The technical staff of the Laboratory of Animal Nutrition and Animal Product Quality and of the Animal Science Unit of the Institute for Agricultural and Fisheries Research involved in this experiment are gratefully thanked for their skilled assistance. None of the authors has any financial or personal conflict of interest to disclose. The study was conducted by own funding. R. P. conducted the laboratory analysis and was involved in data analysis and writing of the manuscript. G. M. T. assisted in data analysis and manuscript writing. G. H. was involved in setting up and carrying out the animal experiment. K. R. and S. D. S. were involved in the experimental setup, data analysis and manuscript writing.
