Whole-body β-oxidation of 18:2ω6 and 18:3ω3 in the pig varies markedly with weaning strategy and dietary 18:3ω3

Segregated early weaning (SEW) into a cleaner nursery increases food intake and growth in pigs, presumably because of reduced immune stimulation compared with conventionally reared, nonsegregated pigs (NSW). The aim of the present study was to evaluate the oxidation of linoleic acid (18:2ω6) and α-linolenic acid (18:3ω3) in SEW and NSW pigs. Pigs consumed a control or high 18:3ω3 diet (ω6 PUFA/ω3 PUFA; 21.3 vs. 2.5, respectively) and were weaned at either 14 days old into a SEW nursery or at 21 days old into a conventional NSW nursery. The major acute-phase protein of pigs but not haptoglobin increased in 35-day-old NSW pigs. NSW pigs had 15–25% lower carcass 18:2ω6 and 20–30% lower carcass 18:3ω3 (% composition) at 49 days old. Between 35- and 49-days-old, NSW pigs had a higher whole-body oxidation of 18:2ω6 (40–120%) and 18:3ω3 (30–80%). The high 18:3ω3 diet decreased the whole-body oxidation of 18:2ω6 by 73% and of 18:3ω3 by 63% in NSW pigs. We conclude that moderately cleaner housing SEW significantly decreases 18:2ω6 and 18:3ω3 oxidation in pigs.

PUFAs are eicosanoid precursors and also contribute importantly to membrane structure. Despite their important precursor role, 18:2 6 and 18:3 3 are primarily ␤ -oxidized for energy (1,2). Although dietary factors like fasting-refeeding and energy restriction are known to significantly increase the ␤ -oxidation of 18:2 6 and 18:3 3, overall little is known about why these two dietarily important fatty acids are mostly ␤ -oxidized.
In this study, we were interested in assessing the effect of weaning strategy and diet on 18:2 6 and 18:3 3 ␤ -oxidation in young growing pigs. Segregated early weaning (SEW) is a management strategy in which pigs are weaned at 10-14 days old instead of the more common 21 days old and are housed in a cleaner environment compared with nonsegregated weaning (NSW) used in conventional commercial units. SEW pigs eat more and grow faster, and there is some evidence that this is due to reduced antigen exposure, i.e., to less demand on their immune system (3)(4)(5)(6)(7)(8)(9). We chose this model to further examine parameters affecting 18:2 6 and 18:3 3 ␤ -oxidation because there are a few, mostly indirect, studies suggesting a stimulated immune system could contribute significantly to the body's oxidation of fatty acids. For instance, during severe inflammation, energy expenditure is increased and the respiratory quotient is decreased, indicating an increase in whole-body fat oxidation (10)(11)(12)(13)(14). The extent to which PUFA such as 18:2 6 contributes to this inflammation-Bazinet et al. Weaning strategy and PUFA ␤ -oxidation in pigs induced increase in fat oxidation is not known. In a rat model of sepsis caused by puncture of the cecum, 14 CO 2 recovery 6 h after an oral dose of [ 14 C]18:2n-6 was shown to be 1.5-2 times higher than in controls (15). As well, rats infused with tumor necrosis factor-␣ (TNF ␣ ) have 10% less 18:2 6 in carcass triglyceride, suggesting that 18:2 6 may be mobilized under these conditions (16). Hence, we postulated that ␤ -oxidation of 18:2 6 and 18:3 3 may decrease in animals undergoing a reduced immune challenge.
The objective of the present study was therefore to assess whole-body PUFA balance in NSW and SEW pigs consuming a conventional compared with an increased level of dietary 18:3 3. Whole-body PUFA balance was chosen as the method to study 18:2 6 and 18:3 3 oxidation because, with this method, ␤ -oxidation is determined in relation to the intake of 18:2 6 or 18:3 3 over several days to weeks, thus giving a long term comparison between PUFA utilization as fuels versus incorporation into tissue membrane structure or as long chain PUFA precursors (1). We hypothesized that SEW pigs would have lower whole-body 18:2 6 and 18:3 3 ␤ -oxidation as a result of less immune stimulation due to being housed in a cleaner environment with lower antigen exposure. In view of the anticipated higher oxidation of 18:3 3 in the NSW pigs, we wanted to assess whether raised 18:3 3 intake would influence 18:2 6 and 18:3 3 balance, especially in the NSW pigs.
In the pig, the plasma acute-phase proteins, haptoglobin, and major acute-phase protein of pigs (pig-MAP), are sensitive indicators of inflammation (17). Therefore, in the present study, haptoglobin and pig-MAP were determined at various times before and after weaning.

Experimental animals and treatments
Pregnant Yorkshire-Landrace sows (Maple Leaf Foods Agresearch, Burford, ON) were randomly fed either a control diet or a diet high in 18:3 3 from ‫ف‬ 1 week prior to parturition until weaning. The diets consisted of base commercial pig feed (Shur-Gain Feed # 694, Guelph, ON) containing 477 g/kg of corn, 270 g/kg of soy bean meal, 75 g/kg of wheat middlings, 74 g/kg of whey powder, 36.2 g/kg of fish meal, 10 g/kg of limestone, 7 g/kg of dicalcium phosphate, 3.3 g/kg of salt, and 17.6 g/kg of a vitamin/mineral supplement. The control diet consisted of the base pig feed supplemented with 50 g/kg of sunflower oil, while the high 18:3 3 diet consisted of the same base pig feed supplemented with 15 g/kg of sunflower oil and 35 g/kg of flaxseed oil. The fatty acid profile of the two diets is shown in Table 1 . The piglets in the present study suckled from their respective sows consuming either the control or high-18:3 3 diet. At 14 days old, 12 suckling pigs per dietary treatment were randomly allocated to either the SEW or NSW treatment. SEW pigs were weaned at 14 days old into a "clean" nursery while NSW pigs were weaned at 21 days old into a conventional nursery. The SEW nursery was made cleaner by segregating pigs in this study from all other pigs except for littermates, while NSW pigs remained in the same nursery as the sow and nonlittermates (3,9). After weaning, pigs consumed diets with the same 18:3 3 content as their respective sows until the end of the study. Pigs were anesthetized and killed at 14, 35, or 49 days old for tissue proximate and fatty acid analysis. Blood samples were anticoagulated with citrate. Organ distribution of fatty acids was evaluated in liver, brain, viscera, and carcass. Liver, brain, viscera, and carcass were completely homogenized four times using a Hobart Grinder with 0.12 cm dye. An 80 g sample of the carcass homogenate was freeze dried for proximate analysis. A 5 g sample of the liver, brain, viscera, and carcass homogenate were stored at Ϫ 20 Њ C for fatty acid analysis. The term "whole-body" refers to the carcass, brain, liver, and viscera combined, while "carcass" refers to wholebody minus the liver, brain, and viscera.

Tissue composition
Carcass dry matter content was determined by weighing before and after 2 h at 100 Њ C in a drying oven. Dried samples were then analyzed for fat, protein, and ash, according to the American Oil Chemists' Society method Ba 3-38 and the Association of Official Analytical Chemists methods 990.03 and 925.23, respectively (18,19).

Lipid extraction and fatty acid analysis
Total lipids of plasma and homogenized samples of liver, carcass, viscera, and brain were extracted into chloroform-methanol (2:1, v/v). Nonesterified heptadecanoic acid (Sigma, St. Louis, MO) was added as an internal standard to an aliquot of the total fatty acid extract to quantitate total lipids. Total lipids were then recovered and saponified in methanolic potassium hydroxide (60 g potassium hydroxide/l-methanol) for 1 h at 100ЊC and the fatty acids were converted to fatty acid methyl esters using 14% boron triflouride-methanol at 100ЊC for 30 min (Sigma) under nitrogen (20). Fatty acid methyl esters were analyzed by gas liquid chromatography using a capillary column ( JandW Scientific DB-23, 30 m ϫ 0.25 mm, ID), in a Hewlett-Packard 5890A gas liquid chromatograph (Palo Alto, CA) with automated sample delivery and injection (Hewlett-Packard 7671A) and peak integration (Hewlett-Packard 3393 inte-grator). Total lipid fatty acids were quantified on the basis of the proportion in each chromatogram of the corresponding internal standard added to each sample.

Acute-phase proteins
Haptoglobin and pigMAP were separated with one-dimensional reducing SDS-PAGE on 5-12% acrylamide gels stained with Coomassie Brilliant Blue. Bands for pig haptoglobin ‫44-24ف(‬ kDa) and pigMAP ‫511ف(‬ kDa) were first identified by immunoblot with goat anti-human haptoglobin (Sigma-Aldrich) and rabbit anti-pigMAP (gift of courtesy of Dr. F. Lampreave, University of Zaragoza, Spain). These bands could be readily distinguished and quantified in Coomassie Blue stained gels, which were used to quantify each protein. The density of the respective bands was determined from digital images (using the public domain NIH Image program developed at the National Institutes of Health and available on the Internet at http://rsb.info.nih.gov/nih-image/) and expressed as integrated density units. Integrated density units were correlated directly with milligrams of stained protein in the gel loading conditions used, as derived from a standard curve of integrated pixel density versus known amounts of BSA as a reference (21).

Whole-body fatty acid oxidation
Whole-body fatty acid oxidation (disappearance) was determined as previously described according to equation 1: disappearance ϭ intake Ϫ (accumulation ϩ excretion). (Eq. 1) Actual fatty acid excretion was not measured, but a reference value of 3% of fatty acid intake was used (1). Food intake and fatty acid accumulation were measured between 35 and 49 days old.

Data analysis
All data are expressed as means Ϯ SD. Statistical comparisons were made using two-way ANOVA with weaning (SEW and NSW) and diet (control and high 18:33) as the main independent variables and sow as a nested variable. Two-way ANOVA was used to test the independent effects of weaning, diet, and weaning by diet interactions. All comparisons were made pair-wise using Tukey's test. Correlations were calculated using Pearson's test. A probability of less than 0.05 was consid-

Feed intake and body weight
SEW pigs consumed 57% more feed between 35 and 49 days old than NSW pigs, a difference reflected in the 15% heavier body weight of SEW pigs at 49 days old (P Ͻ 0.05; Table 2).

Plasma acute-phase proteins
There was no effect of diet on plasma haptoglobin or pigMAP at 14 days old (Fig. 1). At 35 days old, SEW pigs had lower levels of pigMAP (P Ͻ 0.05) but not haptoglobin (P ϭ 0.06). There were no differences in haptoglobin or pigMAP at 49 days old.

Proximate composition
At 35 days old, SEW pig carcasses contained 7% more water, but 28% less fat, 12% less protein, 11% less ash, 15% less calcium, and 15% less phosphorus (P Ͻ 0.05 vs. NSW; data not shown). There were no differences in the proximate composition of 49-day-old pigs.
Plasma and tissue fatty acids. At 49 days old, weaning strategy had no significant effect on plasma total lipid fatty acid profiles. However, NSW and SEW pigs consuming the high 18:33 diet had higher levels of plasma 18:33 and 20:53 but not 22:63, and had lower levels of 18:26 and 20:46 (Table 3). At 49 days old, NSW pigs consuming the control diet (lower 18:33 intake) had higher total fatty acids in liver total lipids ( Table 4). Pigs consuming the high 18:33 diet had higher carcass 18:33 and 20:53 but lower 18:26 and 20:46 relative to pigs consuming the control diet. At 49 days old, SEW pig carcasses contained more 3 and 6 PUFA but less saturated and monounsaturated fatty acids than NSW pigs ( Table 5). As well, SEW pigs consuming the high 18:33 diet had higher 3 PUFA but lower 18:26 in the carcass. Plasma 20:46 was positively correlated with pigMAP (r 2 ϭ ϩ0.46, P Ͻ 0.001) and haptoglobin (r 2 ϭ 0.55, P Ͻ 0.001), while plasma 20:53 was negatively correlated with pigMAP (r 2 ϭ Ϫ0.29, P Ͻ 0.05) and haptoglobin (r 2 ϭ Ϫ0.28, P Ͻ 0.05; data not shown). SEW, segregated early weaning; NSW, nonsegregated weaning; SFA, saturated fatty acids; MUFA, monounsaturated fatty acids; PUFA, polyunsaturated fatty acids; NS, not significant; D, diet effect; W, weaning effect; DXW, diet by weaning interaction. P Ͻ 0.05. a Percentage of total fatty acids Ͼ12C, n ϭ 6 per group, mean Ϯ SD. b Sum of fatty acids includes more than those shown. c Total extracted fatty acids based on liver wet weight.

DISCUSSION
The present study demonstrates that NSW pigs appear to oxidize all the 18:26 and 18:33 they consume after weaning and that SEW markedly reduces 18:26 and 18:33 oxidation. SEW probably contributed to lower whole-body PUFA oxidation in part by being a cleaner environment and presumably by reducing immune stimulation, as indicated by the lower plasma pigMAP in SEW pigs at 35 days old (Fig. 1). The high 18:33 diet also reduced whole-body oxidation of both 18:26 and 18:33 but only in the NSW pigs.
SEW raised feed intake (Table 2), which probably explains the heavier body weights in this group also reported elsewhere (3,5,9). Higher feed intake after weaning may have prolonged effects on body weight, even though the difference in plasma pigMAP was transient and disappeared by the time the pigs were 49 days old. Although not measured in the present study, cytokines such as TNF␣ would be expected to be lower in the SEW com-pared with NSW pigs, which would increase appetite and growth in the former group (22).
The 40% higher total fatty acid content of the liver in NSW pigs on the control (low 18:33) diet is consistent with other animal models of infection (23,24). Higher 18:33 intake significantly reduced total fatty acid accumulation in NSW pig livers, mostly by lowering 6 PUFA, (Table 4), an effect consistent with reports showing that raised intake of 20:53 inhibits fatty acid accumulation in the liver of rats infused with lipopolysaccharide and Escherichia coli (23,24).
SEW pigs had higher carcass levels of 18:26 and 18:33, but lower levels of saturates and monounsaturates (Table 5), reflecting differences in both weaning strategy and 3 PUFA intake. Thus, compared with NSW, SEW preserved carcass PUFA, thereby contributing to markedly reduced whole-body disappearance (oxidation) of 18:26 and 18:33. Perhaps surprisingly, higher 18:33 in NSW pigs helped reduce 18:26 and 18:33 oxidation (Fig. 2). It seems likely that some of the PUFA released from the carcass stores contributed to PUFA accumulation in the liver of the NSW pigs. This appears to be similar to the process described as "triglyceride recycling" in models of infection (25).
One potential mechanism for the lower carcass PUFA in NSW pigs could relate to the action of hormone sensitive lipase. Hormone sensitive lipase is responsible for releasing free fatty acids from adipose triglycerides into the circulation. In vitro studies with adipocytes have shown that when stimulated with stress hormones, hormone sensitive lipase preferentially releases free PUFA into the medium (26).
From the present study it is not clear how higher 18:33 intake modified the effects of SEW on PUFA metabolism. 18:33 did not decrease the acute-phase protein response at any time point measured (Fig. 1). However, plasma 20:46 was positively correlated and 20:53 was negatively correlated with the levels of haptoglobin and pigMAP (see Results). 20:53 has potent anti-inflammatory properties that are thought to arise from its ability to decrease T-cell proliferation as well by competing with 20:46 for eicosanoid production (27,28,29). 20:53 is also a ligand for peroxisome proliferator-activated receptors regulating the transcription of many proteins involved in fat metabolism (30,31). 18:33 is slowly converted to 20:53, which may possibly explain some of the anti-inflammatory properties of 18:33 (32,33).
In summary, SEW resulted in faster growth, less total fat in the liver, higher PUFA in the carcass, and a marked decrease in the whole-body disappearance of both 18:26 and 18:33. By reducing whole-body oxidation of 18:26 and 18:33, dietary 18:33 mimicked some of the metabolic changes in PUFA metabolism induced by SEW. The mechanism by which this occurred is not known. We conclude that moderately cleaner housing after weaning (SEW) markedly reduces PUFA oxidation. This effect on PUFA metabolism is probably linked to less immune stimulation in the SEW compared with