Measuring plasma fatty acid oxidation with intravenous bolus injection of 3H- and 14C-fatty acid.

Accurate measures of plasma FA oxidation can improve our understanding of diseases characterized by impaired FA oxidation. We describe and compare the 24 h time-courses of FA oxidation using bolus injections of [1-(14)C]palmitate versus [9,10-(3)H]palmitate under postabsorptive, postprandial, and walking conditions. Fifty-one men and 95 premenopausal women participated in one condition (postabsorptive, postprandial, or walking), one tracer ((14)C- or (3)H-labeled), and an acetate or palmitate study. Groups were matched for sex, age, and body mass index (BMI). At 24 h, cumulative [(3)H]acetate recovery as (3)H(2)O was 80 ± 6%, 78 ± 2%, and 81 ± 6% in the postabsorptive, postprandial, and walking conditions, respectively (not significant). Model-predicted maximum [1-(14)C]acetate recovery as expired (14)CO(2) was 59 ± 12%, 52 ± 8%, and 65 ± 10% in the postabsorptive, postprandial, and walking condition, respectively (one way ANOVA, P = 0.12). When corrected with the corresponding acetate recovery factors, 24 h time-courses of FFA oxidation were similar between [1-(14)C]palmitate and [9,10-(3)H]palmitate in all three conditions. In contrast to previous meal ingestion studies, an acetate-hydrogen recovery factor was needed to achieve comparable oxidation rates using an intravenous bolus of [(3)H]palmitate. In conclusion, intravenous boluses of [9,10-(3)H]palmitate versus [1-(14)C]palmitate gave similar estimates of 24 h cumulative FFA oxidation in age-, sex- and BMI-matched individuals.

recovery averaged 51% during both rest and physical activity ( 11 ).
When using hydrogen-labeled tracers ( 2 H-or 3 H-labeled FAs), the hydrogen atoms from the oxidation of the tracer appear as 2 H 2 O or 3 H 2 O in body water. During ␤ -oxidation, most of the hydrogen labels ( ‫ف‬ 75%) are transferred to NAD + and FAD ( Fig. 2 ) and, eventually, they appear in body water. The remaining label ( ‫ف‬ 25%) reaches the TCA cycle as part of the methyl group of the acetyl CoA molecule. Because the hydrogens of the methyl group are stereochemically identical, any of three hydrogens can have released as CO 2 after the possibility for sequestration in the glutamate/glutamine pool.
With continuous intravenous infusion studies, the recovery of position-1 carbon-labeled acetate averages 56% in the postabsorptive state, 50% during hyperinsulinemiahyperglycemia, and 80% during exercise ( 7 ). The percent recovery of labeled CO 2 from the oxidation of carbon-labeled acetate is dependent on the duration of the acetate tracer infusion ( 3,4,9 ) as well as the position of the carbon label on the acetate molecule ( 21 ). When position-1 carbon-labeled acetate was ingested with a meal, the carbon   3 H labels of [3, H]palmitoyl-CoA through ␤ -oxidation and two turns of the TCA cycle. The process of 3 H label removal is similar to that for [9,10-the remaining label. Figure 2 presents the fate of each methyl group hydrogen and its likelihood for sequestration in isotopic exchange reactions in the TCA cycle.
The studies that compared hydrogen-and carbonlabeled FAs for fat oxidation measurements were meal experiments; i.e., the acetate tracers were orally ingested ( 13,14 ). These investigators ( 13,14 ) reported that a hydrogenlabeled acetate recovery factor was not required to account for the sequestration of the hydrogen label from FAs in isotopic exchange reactions. No information exists on the use of hydrogen-label FA tracers administered intravenously (continuous infusion or bolus injection) for the assessment of FA oxidation in humans.
The goals of the present study were 1 ) to describe and compare the 24 h time courses of cumulative FA oxidation assessed with a bolus injection of [1-14 C]palmitate versus [9,10-3 H]palmitate under three conditions: postabsorptive, postprandial, and during walking; and 2 ) to provide [1-14 C]

acetate and [
3 H]acetate recovery factors. The data were collected as part of a larger study examining the oxidative and the nonoxidative disposal of plasma FFAs in humans. Data related to the nonoxidative aspect of the study have been published elsewhere ( 17,18 ).

Study participants
After approval from the Mayo Institutional Review Board, a total of 51 men and 95 premenopausal women gave informed written consent to participate in the study. The following criteria were used to select the volunteers: body mass index (BMI) 18-36 kg/m2; no regular, vigorous physical activity; normal urinalysis; complete blood count; electrolytes; liver and kidney function tests; no medications known to infl uence lipid metabolism (including oral contraceptives); and no tobacco. All subjects were weight-stable for at least 2 months before the study.
Each volunteer participated in one condition (postabsorptive, postprandial, or walking), one tracer (14C-or 3H-labeled) and either the acetate or the palmitate study. In the postabsorptive protocol, volunteers remained fasted until 1330 h on Day 1. They received a fat-free lunch at 1330 (70% of energy requirements) and a mixed supper (30% of energy requirements) at 1800. The volunteers rested throughout this protocol until the end of the study at 0800 next day (Day 2) and drank water ad libitum .
In the postprandial protocol, at 0620 h on Day 1, the participants began consuming small portions of a fat-free smoothie [fat-free frozen yogurt, skim milk, Beneprotein (Nestlé Nutrition), Polycose (Abbott Nutrition), and frozen unsweetened strawberries] at 20 min intervals until 1720 h. Overall, the smoothie portions provided 70% of each individual's daily resting energy requirements. The smoothie provided 30% of energy as protein and 70% as carbohydrate. The participants received a mixed supper at 1800, which provided the remaining 30% of daily resting energy requirements. The volunteers rested throughout this protocol until the end of the study at 0800 on Day 2 and drank water ad libitum.
In the walking protocol, volunteers began walking on the treadmill at ‫ف‬ 2 mph at 0700 h on Day 1. They continued walking for 5.5 h, i.e., until 1230 h. As in the postabsorptive protocol, volunteers remained fasted until 1330, and they received a fatfree lunch at 1330 and mixed supper at 1800. Volunteers rested until the end of the study at 0800 on Day 2 and drank water ad libitum.
When the 14 C tracers (acetate or palmitate) were administered, breath samples were obtained at frequent intervals until the next morning at 0800 h for measurement of 14 CO 2 specifi c activity (SA). Breath CO 2 production rates were measured by indirect calorimetry (DeltaTrac Metabolic Cart; Yorba Linda, CA) hourly until 1700 on Day 1 and at 0800 on Day 2. The metabolic cart was calibrated each morning of the study. Additional quality control for the metabolic carts included monthly pressure calibrations and gas calibrations together with every 6 month calibrations using an alcohol burn test.
When the 3 H tracers (acetate or palmitate) were administered, blood samples were obtained at frequent intervals on Day 1 until the next morning at 0800 h for measurement of plasma 3 H 2 O SA (dpm/ml). In addition, urine was collected for 24 h after the tracer bolus to assess 3

Assessment of residual radioactivity in plasma lipids.
In small subsets of the study participants (postabsorptive protocol, n = 7; postprandial protocol, n = 6; walking protocol, n = 6), we assessed the residual radioactivity ( 14 C-and 3 H-) in plasma lipids for 8 days after the palmitate tracer bolus. One milliliter of plasma was extracted with chloroform-methanol 2:1, the solvent was evaporated, and the lipids were counted with scintillation counting. The results (dpm/ml plasma) were multiplied by plasma volume (PV = 55 ml × kg FFM) to calculate the total residual radioactivity in plasma. The results are expressed as percent of tracer dose administered.

Body composition
Fat free mass and body fat mass were measured using dual energy X-ray absorptiometry (DXA, DPX-IQ, Lunar Radiation; Madison, WI).

Participant characteristics
The 24 h time courses of the cumulative percent recovery of [ 14 C]acetate in breath CO 2 is depicted in Fig. 3 (right panel). The recovery of 14  For the measurement of total body water with 2 H 2 O, 500 µl of the urine sample was added to 40 mg activated charcoal in a 12 × 75 mm tube and mixed. After centrifugation, the supernatant was passed through a spin-x fi lter (0.22 µm) and 200 µl transferred to a 300 µl autosampler vial. Deuterium in the 'decolorized' urine was measured using a Thermo Delta V Advantage Isotope Ratio Mass Spectrometer (IR/MS) equipped with a high-temperature carbon-reduction elemental analyzer (TC/EA) inlet. One microliter of urine was reduced to hydrogen-deuterium by injection into the on-line glassy carbon TC/EA reactor held at 1400°C. Gases produced in the reactor were separated on a 5 Ǻ molecular sieve gas solid chromatography column prior to introduction into the IR/MS. Quadruplicate aliquots of each sample were measured against a calibration curve normalized to the standard mean ocean water scale analyzed in the same sequence. Total body water was calculated using the formula by Schoeller et al. ( 22 ).

Calculations
The rate of 14 CO 2 production in breath (dpm/min) was determined by multiplying expired air 14 CO 2 SA (dpm/mmol) by CO 2 production rate (mmol/min) at each time point. A nonlinear model was used to predict nocturnal CO 2 production rates ( 23 ), because nocturnal CO 2 production rates were not measured in this study.

Statistical analysis
Values are expressed as means ± SD, unless otherwise indicated. One-way ANOVA was used to compare the demographic characteristics of the volunteers among the three conditions. Figure 5 depicts the 24 h time courses of the cumulative [9,10-3 H]palmitate and [1-14 C]palmitate oxidation in the postabsorptive, postprandial, and walking protocols. These data were not corrected for carbon or hydrogen loss. Previous studies have found that a hydrogen-labeled acetate recovery factor is not required when measuring dietary fat oxidation using hydrogen-labeled FAs ( 13,14 ). To assess whether this was also true when tracers were administered intravenously, we compared the 24 h and B (in h Ϫ 1 ) is the rate constant k of the increase in percent recovery. Examples of the fi t in a randomly selected volunteer from each condition are presented in Fig. 4 . The predicted maximum percent recovery of [1-14 C]acetate (model parameter A) was 59 ± 12% (postabsorptive), 52 ± 8% (postprandial), and 65 ± 10% (walking) (one way ANOVA P = 0.12). The inter-individual CV of the predicted maximum percent [1-14 C]acetate recovery ranged from 16.9% to 20.3% ( Table 2 ).

Palmitate oxidation
The  Values are means ± SD; BMI, body mass index.  ( Fig. 6 ). In the postabsorptive and walking conditions, Values are means ± SD; CV, coeffi cient of variance. For sample sizes in each protocol see Table 1 .   In all three conditions, the oxidation time courses were virtually identical between the two tracers, when the recoveries of both acetate labels were applied to correct for label loss.

Post-study residual radioactivity in plasma lipids
There was very low residual radioactivity in plasma lipids at 48 h after the tracer bolus (postabsorptive 1.0 ± 0.1%; postprandial 1.4 ± 0.4%; walking 1.0 ± 0.1%). On day 4,  C]palmitate oxidation, respectively, provided the best agreement between the two FA tracers in the postprandial protocol in our study ( Fig. 7 , middle panel).

Applying the [ 3 H]acetate and [1-
14 C]acetate recovery factors to account for label sequestration resulted in the same 3 H-and 14 C-labeled palmitate oxidation rates in the postabsorptive resting and walking conditions. It is wellestablished that carbon-labeled FA tracers need to be corrected for carbon retention and fi xation in the TCA cycle. Therefore, it was not surprising that without acetatecarbon correction, the time course of cumulative [1-14 C] palmitate oxidation was ‫ف‬ 40% lower than that of [9, H] palmitate oxidation ( Fig. 5 ) ( Fig. 6 , middle panel). It is possible that because FA oxidation is suppressed postprandially, the inclusion of the relatively small acetate-hydrogen recovery factor does not have a major  We found that the appearance of 14 CO 2 from [1-14 C]acetate increased at an ‫ف‬ 50% faster rate in the walking protocol than in the other two conditions. Faster (rather than greater) carbon-label recovery during walking is probably due to more-rapid equilibration of the labeled carbon with the bicarbonate/CO 2 pool(s). Notably, [ 14 C]acetate recovery was always much slower than [ 3 H]acetate recovery ( Fig. 3 ), probably because the hydrogen label mixes very rapidly with the body water pool, whereas longer time is required for the carbon label to transit the bicarbonate pools.
The inter-individual variability of acetate-carbon recovery (16.9-20.4%) was greater than previous reports in continuous infusion studies (12.0-16.1%) ( 5 ) or dietary fat oxidation studies (10.6-12.2%) ( 11 ). The somewhat greater inter-individual variance in our studies partly refl ects the fact that we included women and men with a wide range of adiposity, but it might also relate to the fact that the tracer was administered as an intravenous bolus injection. Interestingly, the CVs of acetate-hydrogen (2.0-8.0%) were substantially lower than those of acetate-carbon recovery. This suggests that using hydrogen-labeled tracers to study cumulative FA metabolism may be more favorable than using carbon-labeled tracers, because the metabolism of the hydrogen label is characterized by a lower inherent variability.
Because of the high inter-individual CV that they observed in their continuous-infusion acetate-carbon recovery study, Schrauwen et al. ( 4 ) advised investigators to measure the acetate recovery factor for every individual. The present studies were part of a complex protocol assessing the nonoxidative disposal of plasma FFA into different tissues. Because of the complexity of the studies, it was not feasible to assess an acetate recovery factor for each individual and in a larger sample. However, we took precautions to control for factors that affect acetate-carbon recovery, such as sex, age, and adiposity ( 5 ). Another limitation was that the use of 14 C-and 3 H-labeled FAs was not performed in the same group of individuals, so a paired analysis of agreement of the two tracers could not be performed ( 24 ). Finally, it was not feasible to assess intra-subject variability. Despite of the above limitations, the present study provides physiologically relevant and novel information on the use of two different FA labels administered intravenously under three different nutritional conditions.
In conclusion, [9,10-3 H]palmitate and [1-14 C]palmitate tracers administered as an intravenous bolus injection gave similar estimates of 24 h cumulative plasma FA oxidation in age-, sex-, and BMI-matched individuals under postabsorptive, postprandial, and walking conditions. Although the sequestration of 3 H label was substantially lower than that of 14 C, the need to use an acetate-hydrogen recovery factor was not eliminated. Therefore, each FA tracer had to be corrected with its corresponding acetate recovery factor. Future studies should validate hydrogen-labeled and carbon-labeled FA tracers administered as bolus injections in the same group of individuals. was ‫ف‬ 20% greater than that of [9, H]palmitate oxidation ( Fig. 6 ). Thus, although hydrogen label sequestration is not as extensive as carbon label sequestration, in order to optimally measure FFA oxidation using a [9,10-3 H] palmitate bolus, we did need to correct 3 H recovery in body water with the appropriate hydrogen-acetate recovery factor in the postabsorptive, postprandial, and walking conditions.
As shown in Figs. 1 and 2 , the opportunity for hydrogen label sequestration during FA oxidation is less than that for carbon label. If all of the hydrogen label that is transferred to NAD + and FAD during ␤ -oxidation eventually appears in body water, then only ‫ف‬ 25% of the hydrogen label enters the TCA cycle where it may be sequestered in isotopic exchange reactions. Furthermore, if we assume that the percent acetate-hydrogen label sequestration in the TCA cycle is equal to the acetate-carbon label sequestration, we can estimate that the average acetate-hydrogen recoveries should be ‫ف‬ 90% in the postabsorptive condition [100 −(25% × 41%)], ‫ف‬ 87% in the postprandial condition [100−(25% × 48%)], and ‫ف‬ 91% in the walking condition [100−(25% × 35%)]. In the present study, cumulative [ 3 H]acetate recovery was ‫ف‬ 80% in each condition, which is somewhat lower than the theoretical values. A previous study reported a cumulative [ 2 H]acetate recovery of ‫ف‬ 88%, when acetate was orally administered and volunteers were subjected to low-intensity exercise ( 13 ). It is unknown whether the lower acetate-hydrogen recovery in the present study is related to the different route of tracer administration (intravenous vs. oral), the different population, and/or the different methodologies employed ([ 2 H]-vs. [ 3 H]-labeled tracers).
Investigators using a continuous intravenous tracer infusion found that acetate-carbon recovery was greater during exercise than during rest (7)(8)(9). A nonlinear relationship has also been described between acetatecarbon recovery and oxygen consumption ( 7 ). During exercise, TCA cycle activity is accelerated in relation to the exchange reactions, perhaps reducing the sequestration of acetate-carbon label in exchange reactions. In our experiments, the maximum [1-14 C]acetate recovery in 14 CO 2 was not statistically different among the postabsorptive, postprandial, and walking conditions. However, at the slow walking speed of ‫ف‬ 2 mph (oxygen consumption of ‫ف‬ 8.5 ml/kg/min vs. ‫ف‬ 3.5 ml/kg/min in the resting condition), it is likely that we did not achieve the high level of TCA cycle activity needed to detect greater acetate-carbon recovery than the resting condition. An alternative explanation for the similar carbon-labeled acetate recovery between the resting and the walking conditions is the route of administration (bolus injection vs. continuous infusion) and the long observation period in our study. In line with our fi ndings, dietary carbon-labeled acetate recovery has been shown to be similar ( ‫ف‬ 51%) between resting and physical activity conditions ( 11 ). Collectively these results emphasize the importance of using identical experimental protocols between acetate and FA tracers in oxidation studies.