Linoleate [18:2(n-6)], arachidonate [20:4(n-6)] and docosahexaenoate [22:6(n-3)] are polyunsaturated fatty acids (PUFA) present in mammalian cellular lipids. Using gas chromatography, labeled fatty acids and in vitro model systems, many variables influencing the desaturation and chain elongation of linoleate to arachidonate and -linolenate [18:3(n-3)] to docosahexaenoate have been intensively studied over several decades (Brenner and Peluffo 1966, Crawford et al. 1989, Garcia and Holman 1965, Sheaff et al. 1995, Sprecher 1981). Nevertheless, the extent to which the conversion of shorter-chain to longer-chain PUFA (LC-PUFA) occurs in the whole animal is still unclear, e.g., for every gram of linoleate consumed, the percentage converted to arachidonate and the influence of individual nutrient deficiencies or overall undernutrition on this process are not well known.

Equally, although tracers and in vitro preparations have lead to abundant information about the -oxidation of PUFA (Bjorntorp 1968, Gavino and Gavino 1991, Jones 1994, Leyton et al. 1987), the applicability of these data in the whole animal has not been adequately investigated. Given the importance of arachidonate and docosahexaenoate during fetal and early post-natal development (Bourre et al. 1990, Carlson 1995, Koletzko and Braun 1991, Neuringer et al. 1986, Sinclair and Crawford 1972) and the high energy requirements of developing neonates, especially the neonatal brain, it is important to be able to reliably and (if possible) quantitatively estimate how much linoleate and -linolenate are -oxidized in comparison with their conversion to LC-PUFA. This may be especially important in infants consuming milk formulas containing no LC-PUFA (Agostoni et al. 1995, Farquharson et al. 1995, Makrides et al. 1995).

The whole-body fatty acid balance method has been developed to address these questions in laboratory animals. Without using tracers, this quantitative method can be used to determine the partitioning of dietary linoleate and -linolenate between accumulation, conversion to LC-PUFA, and disappearance or apparent oxidation. Conditions such as pregnancy and specific or general nutritional deprivation reveal that partitioning of linoleate and -linolenate between accumulation, conversion to LC-PUFA, and -oxidation as measured by whole-body balance methodology is responsive to altered nutrient and energy demands or availability (Chen and Cunnane 1993, Cunnane and Yang 1995). Using whole-body fatty acid balance studies, we have shown that pregnant rats consuming a semipurified diet containing 5% of energy as linoleate partition dietary linoleate as follows: about 29% is accumulated as linoleate, 8% is converted to (n-6) LC-PUFA, <2% is excreted in the feces as total (n-6) PUFA, and the remainder (61%) disappears (-oxidized); nonpregnant rats -oxidize about the same amount but do not convert as much linoleate (or -linolenate) to their respective LC-PUFA (Cunnane and Yang 1995). These results were obtained using diets in which the only sources of PUFA was linoleate and -linolenate, and therefore the LC-PUFA accumulation had to be derived from the linoleate and -linolenate consumed or already in the body.

The aim of the present study was twofold: 1) to examine linoleate balance in rats consuming linoleate at a level considered to be the safe lower limit for nutritional adequacy, i.e., 2% of energy (Bourre et al. 1990, Holman 1971, NRC 1980); and 2) to determine the organ partitioning of linoleate and (n-6) LC-PUFA that accumulate during the balance period. Data on the balance and partitioning of -linolenate and (n-3) LC-PUFA were also obtained, and PUFA partitioning into different organ compartments was compared with that of saturated and monounsaturated fatty acids.

Our present results show that growing rats consuming linoleate at about 2% of energy apparently -oxidize 76%, convert 3% to (n-6) LC-PUFA and accumulate 19% as linoleate. Of the amount stored as linoleate, 77% was in fat, mostly in visceral fat, an organ compartment that accounted for only 8% of body weight gain during the study period. These results suggest that in growing rats with a relatively low linoleate intake, most linoleate is -oxidized or stored in fat, presumably for subsequent -oxidation.

MATERIALS AND METHODS

Table 1. Weight changes in organ compartments of growing rats during a 26-d balance period1 [View Table]

The diet ingredients were mixed on site and consisted of the following (g/kg diet): cornstarch (400), casein supplemented with DL-methionine at 400 mg/kg diet (200), sucrose (155), cellulose (100), fat blend (100), AIN-76 mineral mix supplemented with choline at 10 mg/kg diet (100), and AIN-76A vitamin mix (10). The casein, mineral mix and vitamin mix were purchased from Harlan Teklad (Madison, WI); the other ingredients were purchased locally. The fat blend was as follows (g/kg diet): high oleate sunflower oil (50), hydrogenated soybean oil (40), and high -linolenate flaxseed oil (10) providing the following fatty acid composition (g/kg diet): oleate (42.5), stearate (23.2), linoleate (11.5), -linolenate (9.2), palmitate (7.4), and palmitoleate (2.3); the difference to yield 100 g total fat/kg diet is from the glycerol backbone of the triglycerides. The complete diet was stored frozen at 20°C and thawed on a daily basis as required.

The experimental diet was consumed for a 2-wk acclimation period to stabilize the body's fatty acid content to the new semipurified diet so that baseline fatty acid values could be obtained that would reflect the diet to be consumed during the balance period. After this initial period, body weight and food intake were measured and feces were collected over a 26-d balance period. The rats were randomly paired. One rat in each of the six pairs was killed at the beginning of the balance period, and the other was killed at the end of the balance period for organ weight and fatty acid analysis. Rats were anesthetized using carbon dioxide, the spinal cord was severed at the neck, and the following major organ compartments were carefully dissected, washed in saline, blotted and weighed: brain, liver, all skin, gut (empty esophagus to anus), viscera (heart, lungs, kidneys, bladder, pancreas, spleen, testes and accessory reproductive organs), all visible subcutaneous fat, all visceral fat (perirenal, omental, mesenteric, epididymal), and remaining carcass (skeleton, skeletal muscle, cartilage, spinal cord, and sensory organs of the head excluding brain). Hence, excluding blood, the whole body was analyzed. Organ compartments from each rat were stored separately at 20°C in chloroform-methanol (2:1) containing 0.01% butylated hydroxytoluene. Fecal samples were dried at room temperature for 2 d and then stored at 20°C.

Table 2. Linoleate balance during a 26-d period in growing rats1 [View Table]

A known amount of internal standard (heptadecaenoate [17:0]; NuChek-Prep, Elysian, MN) was added to each sample to the equivalent of approximately 15% of the fatty acids, and the large organ compartments (skin, carcass) were carefully diced. The mixture of tissue, internal standard and chloroform-methanol was homogenized using a Brinkmann Polytron; if the samples occupied more than 30 mL, they were divided into separate extraction vials. In each case, the organic phase was separated by the Folch procedure using 9 g/L saline added after homogenization at the equivalent of 25% of the overall extraction volume. Total lipid extraction was performed in triplicate for each sample. The organic phases of each replicate extraction were recombined and transferred to a preweighed round-bottomed flask and the solvent rotary evaporated at 40°C. The total lipid yield was determined gravitimetrically, and the lipid extract was diluted to 5 g/L in chloroform and stored at 20°C under nitrogen. Actual lipid yield was corrected for the internal standard added. The lipid extract was then saponified in methanolic KOH, and the fatty acids in an aliquot of the total lipid extract were methylated at 90°C for 1 h using boron trifluoride in methanol (14 g/L; Sigma Chemical, St. Louis, MO) and analyzed by gas chromatography (Chen and Cunnane 1993). A 30-m capillary column with 0.25-mm i.d. was used (Durabond 23, J&W Scientific, Folsom, CA) in the split mode (1:100) with a temperature program ramping twice from 150 to 210°C. The fatty acid composition of samples of the food was determined by a similar procedure.

Table 3. Organ partitioning of linoleate during a 26-d balance period in growing rats1 [View Table]

Table 4. Organ partitioning of long-chain (n-6) polyunsaturates during a 26-d balance period in growing rats1 [View Table]

For whole-body linoleate or -linolenate balance analysis, the difference between their intake and the sum of their fecal excretion and accumulation (including their respective LC-PUFA) equals their disappearance or apparent -oxidation (Cunnane and Yang 1995). If linoleate and -linolenate are the only PUFA in the diet, then the net accumulation of LC-PUFA during the balance period must be derived from synthesis. The overall accumulation of (n-6) or (n-3) LC-PUFA relative to linoleate or -linolenate intake, respectively, is therefore equal to the net whole-body rate of desaturation/chain elongation excluding losses by -oxidation. Leyton et al. (1987) and Gavino and Gavino (1991) have shown in different model systems that about 20% of [¹⁴C]arachidonate is -oxidized to carbon dioxide in 24 h. Because disappearance of LC-PUFA cannot be measured directly by the whole-body fatty acid balance method, on the basis of these published radiotracer studies, it has been estimated in the present study that an amount equivalent to 20% of the accumulated (n-6) and (n-3) LC-PUFA is -oxidized and disappears from the body (see tables).

Data are expressed as means ± SD. Student's paired t test was used to determine the significance of differences in weights and the percentage of fatty acid composition of each organ compartment at the beginning and end of the study. Fatty acid accumulation was measured using pairs of rats preselected at the start of the study. The body and organ contents of each fatty acid were then compared between the rats in each paired rat killed on d 0 or 26 of the balance period.

RESULTS

Balance analysis over the 26-d study period showed that a total of 8462 mg of linoleate was consumed, 104 mg was excreted, 1580 mg was accumulated as linoleate and 255 mg accumulated as (n-6) LC-PUFA (Table 2). As a proportion of intake, linoleate excretion was 1.2%, accumulation of linoleate itself was 18.7%, and net conversion to (n-6) LC-PUFA (primarily arachidonate) was 3.0%. By difference, linoleate disappearance was 75.5% of that consumed. Of the 1580 mg that accumulated as linoleate, 77% was in the total fat compartment (64% in visceral fat and 13% in subcutaneous fat) and 13% was in the carcass compartment (Table 3). Of the 255 mg that accumulated as (n-6) LC-PUFA, the distribution in organ compartments was similar for skin, visceral fat and carcass (26-29%); however, unlike linoleate, 13% was in liver and no (n-6) LC-PUFA were accumulated in subcutaneous fat (Table 4).

Analysis of -linolenate balance showed that 2.2% of the intake was excreted, 10.9% accumulated as -linolenate and 1.4% was converted to (n-3) LC-PUFA, leaving 84.9% that disappeared (Table 5). Of the 738 mg that accumulated as -linolenate, 79% was in the total fat compartment (67% in visceral fat, 12% in subcutaneous fat) and 10% was in the carcass compartment (Table 6). Of the 95 mg of accumulated (n-3) LC-PUFA, 66% was partitioned towards total fat (35% in subcutaneous fat and 31% in visceral fat), 23% accumulated in liver and 10% in skin (Table 7).

Table 5. -Linolenate balance during a 26-d period in growing rats1 [View Table]

Table 6. Organ partitioning of -linolenate during a 26-d balance period in growing rats1 [View Table]

Table 7. Organ partitioning of long-chain (n-3) polyunsaturates during a 26-d balance period in growing rats1 [View Table]

The organ partitioning of the accumulated saturated and monounsaturated fatty acids was similar to that of linoleate and -linolenate, with about 75% appearing in fat, mostly in visceral fat (Table 8).

Table 8. Organ partitioning of total saturated fatty acids and oleate during a 26-d balance period in growing rats1 [View Table]

DISCUSSION

These results suggest that both linoleate and -linolenate actively participate in energy metabolism, and they support our previous observations using this methodology (Chen and Cunnane 1993, Cunnane and Yang 1995) as well as tracer studies of the -oxidation of PUFA (Bjorntorp 1968, Gavino and Gavino 1991, Jones 1994, Leyton et al. 1987). Disappearance of linoleate or -linolenate may involve recycling of their carbon skeleton through -oxidation to lipids synthesized de novo (Cunnane and Yang 1995), so some of the PUFA carbon that disappeared could have been recycled into lipids synthesized de novo. The present study was not designed to address carbon recycling, so the amount of carbon entering this pathway is unknown.

There are several implications of substantial partitioning of linoleate towards immediate -oxidation and/or storage in fat. First, if only 6.5% of dietary linoleate was accumulated as linoleate or (n-6) LC-PUFA in lean tissue, which accounted for 90% of body weight gain in this study (Table 1), and yet normal growth was attained (average weight gain of 5.5 g/d), then perhaps the absolute linoleate requirement for normal lean tissue growth may be less than the 2% of energy normally recommended (Holman 1971). Holman's extensive evaluation of linoleate requirements in rats was based on diets devoid of -linolenate and oleate [18:1(n-9)], and it seems that a dietary source of -linolenate and oleate may influence the severity of linoleate deficiency (Anderson and Cunnane 1995), which could then lead to a higher estimate of linoleate requirement than might actually be the case. Linoleate provided at 0.4% of energy intake will support normal rat growth and reproduction as well as normal arachidonate levels in all organs except liver (Bourre et al. 1990). Thus, 2% of energy is a safe lower limit for dietary linoleate, but it may well exceed the actual requirement for adequate growth and tissue (n-6) PUFA accumulation in free-feeding animals. Given the important influence of energy intake on linoleate accumulation (Jones et al. 1995), it will be necessary to study linoleate oxidation while energy and linoleate intakes are varied before recommending any change in linoleate requirement.

Second, if the linoleate requirement for growing rats is close to 2% of energy intake, our present data clearly show that a large part of the linoleate consumed was not actually available for lean tissue growth because it was being consumed for energy. This conclusion seems inescapable in view of the high disappearance of linoleate and its partitioning towards storage in fat even though linoleate intake was relatively low, i.e., it is unlikely that the disappearance of linoleate can be accounted for by -oxidation of excessive amounts of dietary linoleate as might arguably be the case in our earlier work, when linoleate was provided at 5% of energy (Cunnane and Yang 1995) or when LC-PUFA were also present in the diet (Chen and Cunnane 1993).

Third, it is unclear from this or other studies whether linoleate stored in fat is potentially available for conversion to (n-6) LC-PUFA. If this is the case, then storage of linoleate in fat, especially visceral fat, may provide some precursor for turnover of the (n-6) LC-PUFA in lean tissue without unnecessarily using linoleate in lean tissue lipids for this purpose.

Fourth, linoleate requirements presumably decrease as adult weight is achieved so that older rats at mature, stable weight would have a lower retention and higher oxidation of PUFA. Hence, healthy, weight-stable human adults probably oxidize virtually all the linoleate they consume except that required to replace (n-6) LC-PUFA and eicosanoids. We are currently exploring the application of whole-body fatty acid balance methodology to human studies and are replacing the organ compartment analysis performed in rats with quantitative MRI and fatty acid analysis based on adipose tissue needle biopsies and lean tissue samples from a human tissue bank.

The whole-body desaturation/chain elongation values we obtained in this fatty acid balance study [3.0% for linoleate conversion to (n-6) LC-PUFA and 1.4% for -linolenate conversion to (n-3) LC-PUFA] have no direct comparison in the literature. Most research on desaturation and chain elongation has used isolated microsomal preparations and radiotracers (Brenner and Peluffo 1966, Sprecher 1981). The whole-body balance method provides an estimate of the net conversion through this pathway in the absence of LC-PUFA in the diet that reflects the overall whole-body capability of rats to produce LC-PUFA under normal conditions of adequate nutrient and energy availability. When nutrients such as zinc are even moderately limiting, this overall conversion is compromised (Cunnane and Yang 1995). Nevertheless, some organs accumulate more LC-PUFA than others, and, in the present study, visceral and subcutaneous fat probably acquired some newly synthesized LC-PUFA from hepatic desaturation and chain elongation.

The diet used in this study provided nearly a 1:1 ratio of linoleate:-linolenate, which is high by most standards but unlikely to have affected the accumulation or partitioning of linoleate (Bourre et al. 1990). We have since completed a similar study but with a dietary linoleate:-linolenate ratio of 6.7:1, and we obtained a similar value for linoleate oxidation (70%; Anderson and Cunnane 1995). Thus, although -linolenate competes with linoleate for desaturation/chain elongation (Garcia and Holman 1965), it seems unlikely that this competition was sufficient to account for the relatively high linoleate oxidation observed here. The low net conversion of -linolenate to (n-3) LC-PUFA (1.4 vs. 3.0% for linoleate, Tables 4 and 7) might have been partly due to the relatively high -linolenate intake, but, overall, organ partitioning of -linolenate was similar to that of linoleate, oleate and saturates (Tables 3, 6 and 8). Eicosapentaenoate [20:5(n-3)] is more readily -oxidized than arachidonate or docosahexaenoate (Gavino and Gavino 1991), and so synthesis of eicosapentaenoate [20:5(n-3)] may also have helped skew (n-3) LC-PUFA accumulation towards fat tissue (Herzberg 1991, Jandacek et al. 1991).

The fatty acid balance method assumes that linoleate or -linolenate disappearance can be calculated from measurements of fatty acid intake, accumulation and excretion. Conversion to eicosanoids and -oxidation of LC-PUFA are two factors that we have not measured, but neither has a dramatic impact on linoleate partitioning. Eicosanoid production in rats as measured by their urinary excretion does not exceed 1 µg/d (Hansen and Jensen 1983). Disappearance of LC-PUFA through -oxidation was estimated from the published literature (Gavino and Gavino 1991, Leyton et al. 1987) as being 20% of intake (or net synthesis and accumulation since LC-PUFA were not consumed in this study). This seems reasonable in relation to measured LC-PUFA excretion at 25-30% of their accumulation. These estimated LC-PUFA oxidation values represent 0.03-0.06% of linoleate and -linolenate intake, respectively, so they could be raised by a factor of 10 and still would have minimal impact on the disappearance data for linoleate or -linolenate. Equally, excretion values for PUFA may overlook conversion of PUFA to non-fatty acid metabolites by gut microbes. While this error has not been accounted for, it seems plausible to regard its magnitude as being no greater than measureable PUFA excretion values which have only a minor effect on the PUFA partitioning data.

In the present study, visceral fat was taken from four distinct locations (perirenal, omental, mesenteric and epididymal) and these sites may not accumulate or metabolize PUFA at the same rate. The visceral site might also accumulate a different proportion of PUFA intake in older compared to growing rats. In the present study, the visceral fat compartment was initially the same size as the subcutaneous fat compartment but, by the end of the study, visceral fat accounted for more of the overall body weight gain than subcutaneous fat by a ratio of 3.7:1 (Table 1). However, accumulation of linoleate and -linolenate in fat favored the visceral compartment by ratios of 4.9:1 and 5.9:1, respectively, suggesting a preference by linoleate and -linolenate for visceral compared to subcutaneous fat (Tables 4, 7).

In humans, visceral fat has an important role in modulating the cardiovascular risks of abdominal obesity (Despres et al. 1990, Ross et al. 1992). The rat model used in this study has no direct relevance to studying the potential importance of visceral fat in obesity, although it does provide a quantitative basis for determining whether visceral fat accumulates a disproportionate amount of some fatty acids or total fatty acids relative to subcutaneous fat. Food deprivation, deprivation-refeeding and weight cycling cause differential mobilization of fatty acids from fat tissue of rats (Chen et al. 1995, Raclot and Groscolas 1995); therefore a future study using fatty acid balance methodology in an animal model of abdominal obesity could address the possible differential accumulation and/or release of fatty acids from these two important fat sites.

We conclude that linoleate and probably -linolenate are largely partitioned towards -oxidation or storage in visceral fat in growing rats. This study raises the possibility that losses of linoleate through -oxidation may contribute to part of the linoleate requirement for normal lean tissue growth.