|
|
||||||||
Original Article |
Correspondence to: Marc K. Hellerstein.
| ABSTRACT |
|---|
|
|
|---|
We used [2-13C1]glycerol to characterize very low density lipoprotein (VLDL)-triglyceride kinetics and intrahepatic glycerol metabolism in normal men (n = 4) after alcohol (EtOH) ingestion. [2-13C1]glycerol was infused before and after the consumption of 48 g EtOH or a placebo. Three additional subjects also received [1-13C1]acetate in addition to the [2-13C1]glycerol with EtOH treatment. Incorporation of tracer into the glycerol or fatty acid moiety of VLDL-triglyceride was measured by gas chromatographymass spectrometry and used to calculate VLDL-triglyceride production rates. Intrahepatic triose-phosphate enrichments were also calculated based on mass isotopomer distribution analysis of plasma glucose. There was no difference in VLDL-triglyceride production rates after 48 g EtOH (11.9 ± 3.7 mg/kg/h) or placebo (14.7 ± 3.3 mg/kg/h). The VLDL-triglyceride rate constants calculated by kinetic modeling using the glycerol and acetate tracers in the combined isotope infusion subjects were very closely correlated (r 2 = 0.94). The peak VLDL-glycerol enrichments after EtOH were 22.5 ± 3.3% versus 7.6 ± 0.8% after placebo (P < 0.001), while intrahepatic triose-phosphate enrichments were 19.8 ± 1.3% and 13.1 ± 1.2% (P < 0.001), respectively. Moreover, the calculated asymptotic VLDL-glycerol enrichments (representing the hepatic
-glycerol phosphate enrichment) were significantly higher after EtOH than placebo. The higher ratio of VLDL-glycerol to triose-phosphate labeling after EtOH suggests a metabolic block at glycerol 3-phosphate dehydrogenase.
We conclude that consumption of 48 g EtOH does not increase VLDL-triglyceride production in normal men but does cause accumulation of tracer in hepatic
-glycerol phosphate.Siler, S. Q., R. A. Neese, E. J. Parks, and M. K. Hellerstein. VLDL-triglyceride production after alcohol ingestion studied using [2-13C1]glycerol. J. Lipid Res. 1998. 39: 23192328.
Supplementary key words: MIDA, alcohol, glycerol phosphate, hypertriglyceridemia, kinetic modeling
| INTRODUCTION |
|---|
|
|
|---|
The consumption of alcohol (or ethanol, EtOH) has been reported to acutely increase serum triglyceride (TG) concentrations in healthy volunteers, particularly when large doses (>50 g) are administered (1) (2) (3) (4) (5). Hypertriglyceridemia has also been observed in some individuals who consume large quantities of EtOH regularly (6) (7). Hypertriglyceridemia is usually associated with an increased risk of coronary heart disease (8), but regular EtOH intake is not associated with increased risk (9) (10). Mechanistic observations on the triglyceridemic response to EtOH may help us to better understand these relationships and the clinical implications of EtOH-induced hypertriglyceridemia.
Increased serum TG levels can result from increased production and/or reduced clearance of very low density lipoprotein (VLDL)-TG. Overproduction of VLDL-TG has been shown to contribute to hypertriglyceridemia syndromes in other settings not associated with EtOH intake (11) (12), but there are some discrepancies in the literature as to whether VLDL-TG production is elevated by EtOH. Some studies have found an increased VLDL-TG production in persons exposed to EtOH (13) (14) (15) (16), while others have not (13) (16) (17). Plasma clearance of TG may also play a role in increasing serum TG after EtOH, as TG clearance has been reported to be either increased (14) (15) (16) or unchanged by EtOH intake (7) (13). Additionally, EtOH has been shown to delay the clearance of dietary TG (13).
The goal of this study was to characterize the kinetic response of serum triglycerides to acute EtOH consumption. We measured VLDL-TG production based on the incorporation of [2-13C1]glycerol in young, healthy volunteers after the consumption of 48 g EtOH or a placebo. In addition, we performed a study in a subset of subjects to compare this method to the incorporation of [1-13C1]acetate into VLDL-TG, as the latter has been used by us (17) (18) and others (19) previously as a technique for the measurement of VLDL-TG production rate. Aspects of glycerol metabolism in plasma and liver are also reported.
| METHODS |
|---|
|
|
|---|
Materials
[2-13C1]glycerol and [1-13C1]acetate were purchased from CIL (Andover, MA) and/or Isotec (Miamisburg, OH). Isotopic purity was greater than 98% for all tracers used.
Subject characteristics
Volunteers (n = 7) were recruited by advertisement and gave informed consent before enrolling in the study. All protocols were approved by the University of California at San Francisco and Berkeley Committees on Human Research. All subjects were males and moderate consumers of EtOH (<120 g/week). None of the subjects had any history of alcoholism, and all subjects had normal lipid profiles and liver enzyme concentrations in serum. None of the subjects had any history of medical diseases, metabolic disorders, or were using any medications with known metabolic effects. Additionally, none of the subjects had diabetes mellitus, nor did they have any family history of diabetes. Subject characteristics are shown in Table 1. Body composition was measured by bioelectrical impedance analysis (BIA, Model #1990B, Valhalla Scientific Inc., San Diego, CA). Body fat and total body water were calculated according to the manufacturer's equations.
|
Study design
Four subjects were admitted twice to the General Clinical Research Center at San Francisco General Hospital for separate studies. During one admission, 4 alcoholic beverages containing 12 g EtOH (40% vodka, Absolut, Ahus, Sweden) mixed with sugar-free lemonade (Kraft General Foods, White Plains, NY) were administered; during the other admission, the sugar-free lemonade was given as a placebo. Each EtOH cocktail contained 84 kcal from EtOH and 4 kcal from the sugar-free lemonade, while the placebo cocktails contained 4 total kcal (as measured by bomb calorimetry). The order of the treatment was randomized, with the second admission following 1 week after the first.
The infusion protocol was as follows. After an evening meal (40% total daily caloric requirement, 55/30/15, carbohydrate/fat/protein) at 1700, a constant intravenous infusion of [2-13C1]-glycerol (0.25 mg/kg LBM/min) was begun at 0500. The infusion continued until 1300, at which time subjects were again allowed access to food. Cocktails (EtOH or placebo) were administered at 0800, 0830, 0900, and 0930. Blood was drawn before tracer administration (baseline), in the hour before the cocktails (pre-EtOH), and half-hourly thereafter.
An additional study was performed in three subjects to validate the method of using [2-13C1]-glycerol incorporation into VLDL-TG as a means to quantify VLDL-TG production. These subjects participated only in the EtOH phase; no placebo phase was adminstered. The infusion paradigm was similar to that described above, with the following modifications. [1-13C1]acetate (1 g/h) was infused at a constant rate from 0400 to 1400 in addition to the [2-13C1]glycerol (0.25 mg/kg LBM/min) that was infused from 0500 to 1300. EtOH cocktails (12 g EtOH in sugar-free lemonade) were administered at 0800, 0830, 0900, and 0930. Blood was drawn before tracer administration (baseline), in the hour before the cocktails (pre-EtOH), and half-hourly thereafter. No other food was consumed during the study.
Metabolite isolation and measurement
Blood alcohol concentrations were measured using a standard kit (Sigma, St. Louis, MO). Plasma TG concentrations were measured enzymatically using a Ciba Corning Express 550 automated analyzer (Chiron, Emeryville, CA).
Plasma glycerol and glucose were isolated by ion-exchange chromatography as described previously (20) (21) (22). Plasma was deproteinized with perchloric acid (1:2), neutralized with 6 N KOH, and loaded with a water wash onto gravity flow columns. One set of columns contained an anion exchange resin (AG 1-X8 Bio-Rad, Hercules, CA) and the other contained a cation exchange resin (AG 50W-X8 Bio-Rad, Hercules, CA). The two columns were used in sequence, with the eluent collected completely and lyophilized. The samples were then divided and derivatized in one of two ways. The tri-acetate derivative was used for the glycerol fraction. One hundred µl of acetic anhydridepyridine 2:1 was allowed to react with the lyophilized sample at room temperature for 15 min. After drying under nitrogen gas, the sample was reconstituted in ethyl acetate for GC-MS analysis. The glucose fraction of the ion exchange column eluent was derivatized to the aldonitrile tetraacetate derivative. Hydroxylamine in pyridine (2%) was reacted with the lyophilized sample for 30 min at 100°C. Twenty-five µl of acetic anhydridepyridine 2:1 was added after cooling and allowed to react for 15 min at room temperature. The solution was then reconstituted in ethyl acetate after drying under nitrogen gas.
VLDL were isolated by ultracentrifugation as described previously (23). The isolated VLDL were extracted according to Folch, Lees, and Sloane Stanley (24). After evaporation to dryness of the extracted VLDL lipids, 1 ml each of chloroform and 3.0 N MeOH-HCl were added and allowed to react at room temperature overnight. The fatty acid methyl esters (FAME) were extracted twice with hexane after the addition of 5% NaCl. The extracted FAME were evaporated to dryness under N2 gas and resuspended in heptane for GC-MS analysis. The aqueous component (containing TG-glycerol) was also evaporated to dryness under N2 gas. This portion was then derivatized as described above for glycerol. This method for VLDL extraction does not isolate TG exclusively; phospholipids and cholesteryl esters are also extracted. Thus, our analyses of VLDL-glycerol and VLDL-palmitate represent input from a mixture of TG, phospholipid, and cholesteryl ester. The composition of each component is as follows (25). Glycerol derived from TG contributes 72.4% of the total glycerol in VLDL, while the remaining 27.6% is present in the phospholipid component. Fatty acids from TG, phospholipids, and cholesteryl ester contribute 74.3%, 18.9%, and 6.8% of the total fatty acids present in VLDL, respectively. As TG contributes the bulk of glycerol and fatty acids to the VLDL particle, we have expressed the changes that we observe in VLDL-glycerol and VLDL-palmitate as representing primarily changes in VLDL-TG.
Mass spectrometry
Gas chromatographymass spectrometry (GC-MS) was performed using an HP 5971 instrument (Hewlett-Packard, Palo Alto, CA) in the selected-ion monitoring mode to determine the isotopic enrichments of glycerol (26), aldonitrile-glucose, and palmitate-methyl ester (26). For glycerol and aldonitrile-glucose, chemical ionization mode was used; palmitate-methyl esters were analyzed with electron-impact ionization. A 10 m DB225 (J&W Scientific, Folsom, CA) column was used for GC-MS analysis of the glycerol-triacetate with the initial temperature 75°C and rising 45°C/min to 160°C, 5°C/min to 171°C, and 40°C/min to a final temperature of 220°C. The M0 and M1 of m/z 159 and 160 were collected. The aldonitrile tetraacetate derivative of glucose was analyzed using a 60 m DB17 GC column (J&W Scientific, Folsom, CA). The initial temperature rose from a starting value of 120°C to a final temperature of 270°C at a rate of 40°C/min; the oven was held at 270°C for 7 min. The ions monitored were m/z 328, 329, and 330 (M0, M1, and M2). A 40 m DB-1 (J&W Scientific, Folsom, CA) column was used for chromatographic separation of palmitate-methyl esters. The initial temperature started at 80°C and rose 40°C/min to 310°C followed by a rise to 325°C at a rate of 3°C/min. The M0 of m/z 270, M1 of 271, and M2 of 272 were collected.
Calculations
The rate of appearance (Ra) of plasma glycerol was calculated by the dilution technique (18) (27), using non-tracer dose of glycerol (see Discussion). The rate of appearance of glycerol was calculated from both pre-EtOH/placebo and post-EtOH/placebo timepoints. Values from the 0740, 0750, and 0800 samples were used for the pre-EtOH/placebo measurements and steady-state values between 09001200 were used for the post-EtOH/placebo measurements. Mass isotopomer distribution analysis (MIDA) was used to calculate the fractional contribution from de novo lipogenesis to the VLDL-palmitate pathway (fractional DNL, references (28) (29) (30)), fractional gluconeogenesis (GNG, reference (22)), and the isotopic enrichment of the precursor of each (hepatic acetyl-CoA and triose-phosphate, respectively, references (22), (28) (29) (30)). Based on the principal of combinatorial probabilities, the enrichment of the true precursor (acetyl-CoA for DNL, triose-phosphate for GNG) was inferred from the ratio of the excess double-labeled to single-labeled species (
A2/
A1) of each molecule. The precursorproduct relationship was then used to calculate the fractional contribution of de novo lipogenesis to the VLDL fatty acids or gluconeogenesis to glucose production. More detailed explanations of the MIDA calculations are presented elsewhere (31). The hepatic triose-phosphate enrichments calculated from plasma glucose by MIDA can be compared to enrichments in VLDL-TG glycerol, as the latter is derived from hepatic
-glycerol phosphate ( Figure 1).
|
VLDL-TG production was calculated by modeling the rise towards plateau in VLDL-glycerol enrichment from [2-13C1] glycerol and/or in newly synthesized VLDL-palmitate from [1-13C1]-acetate incorporation (fractional DNL, 1719). VLDL-glycerol enrichment values from half-hourly samples between 800 and 1300 and fractional DNL values from half-hourly samples from 800 to 1400 were plotted, and the best fit for the curve of each versus time was calculated using the Sigma Plot computer program (Jandel Scientific, San Rafael, CA). The data were fit to the equation y = Ay · (1-e-Ks(t-c) where y = VLDL-glycerol enrichment or fractional DNL, A y = the plateau value of TG-glycerol or fractional DNL, K s (hr-1) = the rate constant of rise to plateau in VLDL-TG, t = time in hours, and c = the lag period before isotope incorporation into secreted VLDL-TG. VLDL-TG production was then calculated as Ks (hr-1) x pool size (mg), where pool size = average TG concentration (mg/dl) x estimated plasma volume (5% of total body weight). VLDL-TG half life was calculated by dividing 0.69 by Ks .
Statistics
All results are presented as means ± standard deviation. Two-factor repeated measures analysis of variance (ANOVA) with time and treatment as the two trial factors were performed for most parameters. Statistical differences in VLDL-TG production rates, Ks, and Ay were calculated using a matched, paired t -test. None of the values for the three combined isotope infusion study subjects were included in any of the statistical analyses except for DNL, which was measured only in this study. DNL and EtOH concentrations were analyzed only for a time effect. Tukey's post-hoc test was used to determine statistical significance (P < 0.05) for all parameters analyzed by ANOVA.
| RESULTS |
|---|
|
|
|---|
Plasma concentrations of metabolites
Blood alcohol concentrations (BAC) rose from undetectable levels at 0800 to a peak of 14.5 ± 3.6 mM at 1000 ( Figure 2). BAC decayed linearly thereafter and was 6.6 ± 2.6 mM at the end of the study (1300). Plasma TG concentrations rose 43 ± 21% (P < 0.01) after EtOH, from a pre-EtOH mean of 112 ± 78 mg/dl to 150 ± 78 mg/dl at 1300 ( Figure 3). The average value over the course of the study was 118 ± 67 mg/dl ( Table 2). Plasma TG concentrations fell 2 ± 18% in the placebo treatment (119 ± 38 mg/dl pre-placebo, 115 ± 42 mg/dl at 1300); the average value over time was 111 ± 31 mg/dl (Table 2).
|
|
|
Plasma glycerol enrichment and flux
Plasma glycerol enrichments (%) showed a trend towards an increase after EtOH and decrease after the placebo ( Table 3), although none of the differences reached statistical significance. The mean plasma glycerol flux (the rate of appearance of glycerol) decreased 25% (not significant, P = 0.07) after EtOH from a baseline value of 2.64 ± 0.72 µmol/kg per min to 1.98 ± 0.87 µmol/kg per min ( Table 4). In contrast, the rate of appearance of glycerol increased 17% (P = 0.07) from a pre-placebo value of 2.54 ± 0.50 µmol/kg per min to 2.96 ± 0.51 µmol/kg per min following consumption of the placebo.
|
|
VLDL-glycerol and hepatic triose- phosphate enrichment
VLDL-glycerol enrichments rose over time in both the placebo and EtOH studies ( Figure 4, Table 3). The initial measurement of VLDL-glycerol enrichment at 0740 in the placebo study was 2.3 ± 1.2%; the final measurements between 1130 and 1300 were 6.8 ± 1.0% to 7.6 ± 0.8%. The 0740 VLDL-glycerol value was similar in the EtOH study (2.9 ± 1.0%), but the 1130 to 1300 values (18.9 ± 2.9% to 22.5 ± 3.3%) were very different (P < 0.001). Thus, EtOH initiated a dramatic and statistically significant increase in VLDL-glycerol enrichments as compared to both pre-EtOH and placebo values (Figure 4).
|
Hepatic triose-phosphate enrichments calculated by MIDA from plasma glucose also increased after EtOH (Table 3 and Table 5). The pre-EtOH value was 12.2 ± 1.4% and increased after EtOH to 19.8 ± 1.3% at 1300. Triose-phosphate enrichments were unchanged from baseline after the placebo, the average value being 12.7 ± 1.1%. Post-EtOH triose-phosphate values from 0900 to 1300 were significantly different than the placebo and the pre-EtOH values.
|
Fractional de novo lipogenesis
Fractional DNL was measured in the three subjects who participated in the combined label portion of this study ( Figure 5, Table 6). The average baseline value was 10 ± 10% and it rose to a final value of 37 ± 7% at 1400 (P < 0.001). One of the three subjects had an unusually high baseline value (20%), adding to the large standard deviation observed for the baseline values. All three fractional DNL curves were modeled to calculate kinetics of VLDL-TG. The DNL precursor enrichment values fell after the consumption of the EtOH; the initial 0800 value was 0.092 ± 0.013 and the steady-state post-EtOH value was 0.041 ± 0.004.
|
|
Kinetic modeling and calculation of triglyceride fluxes
Modeling the rate of the rise toward plateau of VLDL-glycerol enrichment in the EtOH and placebo studies enabled us to calculate several VLDL-TG kinetic parameters (Table 2 and Table 7). The mean Ks was 0.21 ± 0.07 hr-1 in the EtOH study and 0.28 ± 0.08 hr-1 in the placebo study (Table 2). Thus, the half-life of VLDL-TG was 3.6 ± 1.4 h in the EtOH study and 2.7 ± 0.9 h in the placebo study (Table 2). There were no statistically significant differences between the two treatments for Ks or t1/2. The predicted mean asymptotic VLDL-glycerol enrichment (Ay) from the curve-fitting was 41.0 ± 4.0% in the EtOH study and 9.2 ± 1.4% in the placebo study (P < 0.01, Figure 4).
|
In the three subjects participating in the methodological comparison phase of the study, Ks was calculated after EtOH consumption by independently modeling the incorporation of [1-13C1]acetate and [2-13C1]glycerol into VLDL-TG (Table 7). For [1-13C1]acetate, the rate of rise toward plateau of label incorporation into newly synthesized palmitate molecules (fractional DNL) was modeled. The average K s value calculated by the modeling of [1-13C1]acetate incorporation into newly synthesized VLDL-TG palmitate was 0.20 ± 0.08 h-1 and by the modeling of [2-13C1]glycerol incorporation into VLDL-TG was 0.20 ± 0.07 h-1.
The mean VLDL-TG production rate was 11.9 ± 3.7 mg/kg per h in the EtOH study and 14.7 ± 3.3 mg/kg per h in the placebo study, using the [2-13C1]glycerol results. Individual data are listed in Table 2. There were no statistically significant differences between the two treatments. When VLDL-TG production was calculated using the [1-13C1]acetate incorporation into newly synthesized palmitate in the EtOH study, the value of 11.0 ± 3.3 mg/kg per h was almost identical to that calculated with [2-13C1]glycerol.
| DISCUSSION |
|---|
|
|
|---|
The primary goal of this study was to characterize VLDL-TG production rates after the acute consumption of 48 g EtOH in normal volunteers using a new stable isotope method (incorporation of [2-13C1] glycerol into the glycerol moiety of VLDL-TG). Based on this method, there was no difference in VLDL-TG production rates after the consumption of EtOH or a placebo under overnight-fasted conditions. We also performed a pilot study comparing a previously described stable isotope technique for assessing VLDL-TG kinetics (incorporation of [1-13C1]acetate in VLDL-TG palmitate via the DNL pathway, references (17) (18) (19)) to validate the [2-13C1]glycerol method. The two methods gave similar values for kinetic parameters. The consumption of EtOH also induced changes in glycerol metabolism, when compared with the placebo.
The consumption of 48 g EtOH elicited a modest (43%) but significant increase in plasma TG concentrations over 5 h compared to values after the placebo was consumed. Similar increases have been observed by others when similar doses of EtOH were administered (3). There may be a link between EtOH dose and plasma TG response, as other studies in which 2- to 3-times more EtOH was administered over 812 h reported increases in plasma TG concentrations of 6575% (2) (4).
In the absence of dietary TG input, plasma TG concentrations are determined by rates of hepatic VLDL-TG production and the efficiency of TG clearance from the plasma. In the current study, we did not observe any significant differences in VLDL-TG production when 48 g EtOH or a placebo was consumed (Table 2). We have previously reported VLDL-TG production rates of 7.5 mg/kg per h after the consumption of 24 g EtOH (17) but had no control placebo protocol for comparison. In the present study, we observed slightly higher values (11.9 mg/kg per h EtOH, 14.7 mg/kg per h placebo). The rates of VLDL-TG production that we measured in the present study are similar to the 1214 mg/kg per h measured in normal overnight-fasted subjects by others (12) (32) and are lower than those reported for hypertriglyceridemic patients (2024 mg/kg per h, references (11), (12)). As plasma TG concentrations increased 43% but VLDL-TG production rates did not change after EtOH consumption, reduced VLDL-TG clearance must account for elevations in plasma TG when EtOH was consumed (Table 2).
One technical issue that deserves comment relates to the measurement of VLDL-TG kinetics in the non-steady state after EtOH consumption. A non-steady state in plasma TG concentrations was observed after EtOH (Figure 3), and is unavoidable experimentally here, unless one maintained high plasma EtOH concentrations for 1216 h prior to beginning the isotope infusion; this, of course, would be difficult to justify in human subjects on ethical grounds. Formal kinetic analysis demonstrates that the metabolic cause of the non-steady state (increased production rate or reduced clearance efficiency) can nevertheless be established from an isotope labeling experiment, as was performed here. A brief mathematical discussion of this point follows (see ref. (33) for a more extensive discussion).
The accumulation of label in a product during an endogenous labeling experiment is described by the following general equation:

where B* = isotope enrichment of product B, A* = isotope enrichment of precursor A, kd = fraction of B removed per unit time (d-1), and ks = fraction of B replaced by synthesis from A per unit time (= production rate/pool size, units d-1). If a metabolic steady state is present, ks = kd = k, and,

This is the standard kinetic equation (in its differential form) and results in straightforward experimental labeling results in the steady state: reduced clearance (lower k) obviously results in slower accumulation of label in B. Indeed, if A* is known, dB*/dt is directly proportional to k, so that the rate of label accumulation directly reflects k.
If a non-steady state is present, ks
kd. An increase in the pool size must reflect either A) higher production rate, or B) lower clearance efficiency. If situation A) exists (production is higher and clearance is unchanged), the `ks' is by definition higher early, before steady state is reached (ks = production rate/pool size, and pool size has not yet reached its steady state value). Accumulation of labeled B* will therefore be accelerated since ks x A* in equation 1 is higher while kd x B* lags until B* is highly labeled. Accordingly, increased production will indeed result experimentally in a more rapidly labeled product and can be interpreted in this manner.
Alternatively, if situation B) exists (clearance is reduced and production is unchanged), then `ks' is by definition lower from the beginning and up to the new steady state (ks = production/pool size, and pool size increases). Accumulation of labeled B* will therefore tend to be somewhat slower, as ks x A* in equation 1 is lower, while kd x B* does not have an impact until B* becomes significantly labeled. So, reduced clearance will result experimentally in a less rapidly labeled product and can be interpreted accordingly. The take-home message is that directional changes in label accumulation rate in B* will indeed reveal the kinetic mechanism responsible for the change in pool size (production or clearance), in at least a semi-quantitative manner.
Several investigators have measured VLDL-TG production rates in response to acute (13) (14) (16) or chronic (15) (34) EtOH exposure. When VLDL-TG production was measured in response to acute EtOH consumption in normal volunteers, increased rates were observed by some (14) (16) but not by others (13) (16). In a study by Crouse and Grundy (13), normal subjects did not have any differences in VLDL-TG production rates when EtOH was included in fat-free meals; EtOH induced a 45% increase in several obese subjects, however. Wolfe et al. (16) reported that EtOH increased VLDL-TG production above pre-EtOH levels when normal subjects had fasted 69 h but not after a 15-h fast. The increased VLDL-TG production in the 69-h fasted subjects was apparently linked with increased peripheral release of FFA, as both FFA flux and FFA concentrations were increased over baseline levels following EtOH infusion in the 69-h fasted group but not the 15-h fasted group. It is possible that the stress induced by the induction of hypoglycemia stimulated a catecholamine response and subsequently overcame the usual anti-lipolytic effect of acetate (35) (36) (37) (38) and increased peripheral lipolysis (39) in the 69-h fasted subjects. The increased VLDL-TG production after EtOH observed by Wolfe et al. (16) in 69-h fasted individuals and by Crouse and Grundy (13) in obese subjects may have been driven by increased lipolytic rates.
Lipolysis was inhibited non-significantly by EtOH consumption in the present study, as measured by the rate of appearance of glycerol. The rate of appearance of glycerol was assessed by quantifying enrichments (Table 4) of plasma glycerol. While the quantity of [2-13C1] glycerol administered gave enrichments above what could be considered a "tracer" dose, we have elsewhere found no effect on calculated lipolytic rates when adding large doses of [2-13C1] glycerol to tracer doses of [2H5] glycerol (A. Basinger, R. A. Neese, M. K. Hellerstein, unpublished data). After EtOH consumption, we observed a 24% decrease in plasma glycerol flux (Table 4); this is similar to the decrease we observed after the consumption of 24 g EtOH previously (17). While we did not measure FFA flux in the current study, FFA flux has been demonstrated by us (17) and others (14) (16) (40) to be reduced after moderate EtOH consumption. Thus, hepatic VLDL-TG production after EtOH was likely kept in check by the suppression of peripheral lipolysis.
The modeling of [2-13C1] glycerol incorporation into VLDL-TG after EtOH intake (Figure 4) is based on the assumption of a relatively stable enrichment in the intracellular precursor pool,
-glycerol phosphate. If intrahepatic
-glycerol phosphate enrichments are not constant during the post-EtOH period (08001300), fitting the data to a simple exponential that approaches an asymptotic value would not be appropriate. If, for example,
-glycerol phosphate enrichments slowly increased during the hours after EtOH consumption, using the asymptotic VLDL-glycerol value to represent the intracellular precursor pool enrichment during the entire period would result in an overestimation of precursor enrichment and an under-estimation of true Ks (27). Because effects of EtOH on the hepatic redox state occur rapidly (41), however, the enrichments of the intrahepatic
-glycerol phosphate and triose-phosphate pools should rapidly achieve their steady-state values. Comparison of the incorporation kinetics of [2-13C1]glycerol and [1-13C1]acetate into VLDL-TG strongly supports this assumption; the nearly identical Ks and VLDL-TG production rates that were calculated by these independent methods in the same individuals (Table 7) make it unlikely that the glycerol labeling method is subjected to a changing intracellular precursor enrichment.
The maximal VLDL-glycerol enrichments after EtOH were several times higher than the highest measured values after the placebo (Table 3, Figure 4). Peak VLDL-glycerol enrichments reflect intrahepatic
-glycerol phosphate labeling and any residual dilution from unlabeled TG in VLDL. Thus, EtOH either alters the intrahepatic
-glycerol phosphate pool enrichment or the contribution of newly assembled TG to circulating VLDL. In animal systems, intrahepatic
-glycerol phosphate concentrations increase when EtOH is given (42) (43) (44). The biochemical mechanism involves the altered hepatic redox state associated with EtOH oxidation (45) (46). Nicotinamide adenine dinucleotide (NAD) is required as a co-substrate (Figure 1) for conversion of
-glycerol phosphate to dihydroxyacetone phosphate (DHAP) via the enzyme glycerol 3-phosphate dehydrogenase (G3PDH). The obligatory oxidation of EtOH by alcohol dehydrogenase (Figure 1) limits the availability of NAD; our results suggest that the rate of conversion of
-glycerol phosphate to DHAP may thereby be reduced.
These results also provide indirect evidence for reduced flux of
-glycerol phosphate to DHAP during hepatic EtOH oxidation. Under fasting conditions (as with the placebo protocol), the VLDL-glycerol enrichments reached only about 60% of the calculated intrahepatic triose phosphate enrichments (Table 3). The extrapolated plateau VLDL-glycerol enrichments (A0 = 9.2%) from modeling of the incorporation curve were reasonably close to the triose-phosphate enrichments calculated by MIDA of plasma glucose (12.7%). In contrast, both the VLDL-glycerol and hepatic triose-phosphate enrichments were higher when EtOH was given; the VLDL-glycerol enrichments rose to levels greater than the hepatic triose-phosphate enrichment. Extrapolated maximum values of VLDL-TG glycerol (Ay = 41.0%) were even higher post-EtOH and were almost 2-fold greater than hepatic triose-phosphate enrichments (Table 3). These results suggest an accumulation of labeled glycerol at G3PDH (Figure 1), i.e., an isotopic gradient between hepatic
-glycerol phosphate and DHAP.
It is interesting to note that the highest VLDL-glycerol enrichments that we observed post-EtOH were still at most one-half to one-third of the plasma glycerol enrichments (Table 3), and even the asymptotic enrichments predicted by kinetic modeling of the VLDL-glycerol curves were still <65% of the plasma glycerol enrichment. Hepatic
-glycerol phosphate must therefore derive at least in part from other sources besides plasma glycerol. Reshef, Hanson, and Ballard (47) have discussed "glycero-neogenesis," or the synthesis of
-glycerol phosphate from gluconeogenic precursors. We did not determine the metabolic source responsible for the dilution between plasma glycerol and
-glycerol phosphate, but studies in cell systems (48) suggest that flux from DHAP to
-glycerol phosphate may have contributed.
In conclusion, consumption of 48 g EtOH increases plasma TG concentrations but does not increase VLDL-TG production rates in normal healthy men when measured by modeling of [2-13C1]glycerol incorporation into VLDL-glycerol. [2-13C1]glycerol incorporation into VLDL-glycerol gave kinetic results similar to concurrent [1-13C1] acetate incorporation into the palmitate moiety of VLDL. Finally, EtOH causes an accumulation of labeled glycerol in the hepatic
-glycerol phosphate pool, which is likely mediated by hepatic redox changes associated with EtOH oxidation.
| ACKNOWLEDGMENTS |
|---|
This work was supported in part by NIAAA grant 1 RO3 AA10693-01 and NIH grant RR-00083 from the Division of Research Resources for the General Clinical Research Center. We gratefully acknowledge Drs. Mark Christiansen, Peter Linfoot, Mark Hudes, and Lisa Lim for their contributions to this study. We also acknowledge the nurses in the San Francisco General Hospital GCRC for their help.
Manuscript received November 3, 1997; and in revised form May 27, 1998; and in revised form August 18, 1998.
Abbreviations:
EtOH, ethanol; TG, triglyceride; VLDL, very low density lipoprotein; BIA, bioelectrical impedance analysis; LBM, lean body mass; GC-MS, gas chromatographymass spectrometry; MeOH-HCl, methanolic hydrochloric acid; FAME, fatty acid methyl esters; M0, M+0 isotopomer; M1, M+1 isotopomer; M2, M+2 isotopomer; MIDA, mass isotopomer distribution analysis; Ra, rate of appearance; DNL, de novo lipogenesis; GNG, gluconeogenesis;
A1, change in fractional abundance of M+1 isotopomer;
A2, change in fractional abundance of M+2 isotopomer; Ks , rate constant of rise to plateau; Ay, plateau value of glycerol enrichment or fractional DNL based on kinetic modeling; ANOVA, analysis of variance; BAC, blood alcohol concentration; K d, rate constant of removal; FFA, free fatty acids; NADH, reduced nicotinamide adenine dinucleotide; NAD, nicotinamide adenine dinucleotide; DHAP, dihydroxy acetone phosphate; G3PDH, glycerol 3-phosphate dehydrogenase; BMI, body mass index; HDL, high density lipoprotein; t1/2, fractional half-life; MPE, mole percent excess; p, precursor enrichment; Glc, glucose; GAP, glyceraldehyde 3-phosphate; Gly-3P, glycerol 3-phosphate; Gly, glycerol; AcCoA, acetyl-coenzyme A, FA-CoA, fatty-acyl-CoA; Ac, acetate; FA, fatty acid; AcHO, acetaldehyde
| REFERENCES |
|---|
|
|
|---|
-glycerophosphate and triglyceride synthesis in the production of fatty liver by ethanol. Proc. Soc. Exp. Biol. Med. 113:814-817.
This article has been cited by other articles:
![]() |
E. J. Parks and M. K. Hellerstein Thematic review series: Patient-Oriented Research. Recent advances in liver triacylglycerol and fatty acid metabolism using stable isotope labeling techniques J. Lipid Res., August 1, 2006; 47(8): 1651 - 1660. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. J. Murphy Stable isotope methods for the in vivo measurement of lipogenesis and triglyceride metabolism J Anim Sci, April 1, 2006; 84(13_suppl): E94 - E. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. L. Chen, E. Peacock, W. Samady, S. M. Turner, R. A. Neese, M. K. Hellerstein, and E. J. Murphy Physiologic and Pharmacologic Factors Influencing Glyceroneogenic Contribution to Triacylglyceride Glycerol Measured by Mass Isotopomer Distribution Analysis J. Biol. Chem., July 8, 2005; 280(27): 25396 - 25402. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. M. Turner, E. J. Murphy, R. A. Neese, F. Antelo, T. Thomas, A. Agarwal, C. Go, and M. K. Hellerstein Measurement of TG synthesis and turnover in vivo by 2H2O incorporation into the glycerol moiety and application of MIDA Am J Physiol Endocrinol Metab, October 1, 2003; 285(4): E790 - E803. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. M. Burke, G. R. Collier, E. M. Broad, P. G. Davis, D. T. Martin, A. J. Sanigorski, and M. Hargreaves Effect of alcohol intake on muscle glycogen storage after prolonged exercise J Appl Physiol, September 1, 2003; 95(3): 983 - 990. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. K. Trimmer, J.-M. Schwarz, G. A. Casazza, M. A. Horning, N. Rodriguez, and G. A. Brooks Measurement of gluconeogenesis in exercising men by mass isotopomer distribution analysis J Appl Physiol, July 1, 2002; 93(1): 233 - 241. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. W. Patterson, B. Mittendorfer, N. Elias, R. Satyanarayana, and S. Klein Use of stable isotopically labeled tracers to measure very low density lipoprotein-triglyceride turnover J. Lipid Res., February 1, 2002; 43(2): 223 - 233. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. D. Jensen, V. Chandramouli, W. C. Schumann, K. Ekberg, S. F. Previs, S. Gupta, and B. R. Landau Sources of blood glycerol during fasting Am J Physiol Endocrinol Metab, November 1, 2001; 281(5): E998 - E1004. [Abstract] [Full Text] [PDF] |
||||
![]() |
W. Wang, A. Basinger, R. A. Neese, B. Shane, S.-A Myong, M. Christiansen, and M. K. Hellerstein Effect of nicotinic acid administration on hepatic very low density lipoprotein-triglyceride production Am J Physiol Endocrinol Metab, March 1, 2001; 280(3): E540 - E547. [Abstract] [Full Text] [PDF] |
||||
![]() |
W. Wang, A. Basinger, R. A. Neese, M. Christiansen, and M. K. Hellerstein Effects of nicotinic acid on fatty acid kinetics, fuel selection, and pathways of glucose production in women Am J Phys |