J. Anim. Sci. 2006. 84:E94-E104
© 2006 American Society of Animal Science
Stable isotope methods for the in vivo measurement of lipogenesis and triglyceride metabolism1,2
E. J. Murphy3
Department of Medicine, University of California–San Francisco, San Francisco 94110; and KineMed, Inc., Emeryville, CA 94608
 |
Abstract
|
|---|
Synthesis of fatty acids (via de novo lipogenesis) and triglycerides are important factors in fat accumulation and the efficiency of animal production. Recently, new stable isotope methods using heavy water (2H2O) have made possible the safe, and relatively easy, measurement of both of these processes in vivo in animals and humans over prolonged periods. These methods also provide information on the relative contribution of glycolysis and glyceroneogenesis to triglyceride synthesis under different physiological settings. The data suggest that numerous dietary factors, including nutrient composition and caloric content, may affect de novo lipogenesis. Significant differences in de novo lipogenesis have also been seen across species and in different tissues. The rates of triglyceride synthesis have been shown to be affected by diet and to differ significantly between different adipose depots, with metabolically active depots (e.g., visceral fat) having much more rapid triglyceride turnover than subcutaneous depots. Dietary fat and the peroxisome proliferator-activated-
agonist rosiglitazone have both been shown to influence triglyceride synthesis rates and to increase glyceroneogenesis. A significant portion of triglyceride synthesis is not related to triglyceride accumulation but rather is secondary to active lipolysis and reesterification. The application of these new techniques to animals other than rodents will undoubtedly enhance our understanding of adipose tissue biology and could lead to new methods for improving animal production.
Key Words: de novo lipogenesis • deuterated water • glyceroneogenesis • kinetics • stable isotopes • triglyceride
|
INTRODUCTION
|
|---|
Triglycerides (TG) serve a unique role in energy metabolism. In adipose tissue, TG provide a virtually unlimited capacity for energy storage, and this ability to store excess energy has implications both for the current obesity epidemic in humans (Popkin, 2004
) and for efficient animal production. Efficient energy storage in adipose tissue in cows is essential for efficient lactation and milk production. However, excess adipose can result in significant waste in energy use for meat production (Bauman, 1976) and yet fat deposition in the muscle, both in the myocytes and around them (marbling) can improve meat quality (Marshall, 1994). Clearly, a thorough understanding of TG metabolism is crucial to optimizing energy-efficient animal production as well as stemming the tide of obesity in humans. Recently developed stable isotope methods for studying TG metabolism now make possible studies of multiple aspects of adipose tissue kinetics.
Historically, investigations of TG metabolism have focused on the synthesis of fatty acids, often termed lipogenesis or more specifically, de novo lipogenesis (DNL). This focus comes, in part, from a desire to understand the practical relevance of the conversion of dietary carbohydrates into fatty acids, and much has been learned about the factors affecting DNL in humans and animals. However, equally important may be the actual synthesis of the TG molecule itself via generation of 
-glycerol phosphate (-GP) followed by sequential acylation. New data suggest that this TG assembly is an extremely active process, with much of the synthesis accounted for by lipolysis and reesterification of fatty acids, in what could be termed futile cycling. The importance of these processes has been made apparent by the recognition of the crucial role of diacylglycerol acyl-transferase in regulation of TG synthesis and obesity (Smith et al., 2000; Chen and Farese, 2005)
This review will discuss currently practiced stable isotope methodologies for the in vivo measurement of DNL, TG synthesis, and glyceroneogenesis with a focus on the use of deuterated water (2H2O) as a tracer. The use of stable isotopes for other aspects of fatty acid metabolism will be briefly mentioned. An in-depth description of experimental detail is beyond the scope of this review but is provided in the cited literature. Important insights into adipose kinetics gained from the use of these techniques will be reviewed, and a brief discussion of future applications in animal production will be presented.
|
THE USE OF 2H2O AS A METABOLIC TRACER
|
|---|
The use of 2H2O as a metabolic tracer has become increasingly popular in recent years due to its numerous advantages over other tracers. In addition to being relatively inexpensive, 2H2O quickly equilibrates with the total body water (TBW) pool, is easily administered over long periods (necessary for adipose TG and fatty acid kinetics), requires no intravenous infusion, and allows for the simultaneous measurement of several processes. As typically used, animals are given an initial bolus dose of 2H2O in normal saline via intraperitoneal injection, although an oral bolus can also be used. Animals are then maintained on drinking water enriched anywhere from 2 to 10% with 2H2O. The final body water enrichment observed is significantly lower than that of the drinking water, presumably due to dilution from unlabeled metabolic water generated from food and due to water exchange through respiration (Lee et al., 1994a). Table 1 shows final body water enrichments obtained in rodents drinking 2H2O enriched water. As shown, final TBW enrichments range from 56 to 75% of the administered water enrichment in these small animals, with a larger dilution suggested in mice than in rats. There is a physiologic basis for these findings given the greater metabolic rate of smaller animals. Larger animals such as pigs and cows may therefore exhibit significantly less dilution. Correlates with data obtained in other large mammals (i.e., humans) are difficult as typically they are given a fixed volume per day of heavy water (e.g., 40 to 80 mL) and variations in daily body water turnover secondary to differences in exercise and volitional water intake make definitive determination of the actual percentage enrichment administered difficult. Because there is no definitive proof as to what the actual sources of unlabeled water are, it will not be possible to tell a priori what final body water enrichments will be achieved for a given drinking water enrichment in larger animals. This, however, is easily determined empirically.
The goal of the 2H2O bolus is to reach the intended final body water enrichment rapidly. The bolus dose is calculated from estimates of the TBW pool size and intended final body water enrichment. Estimates used for TBW pool size in rodents have ranged from 60 to 70% of body weight (Lee et al., 1994a; Brunengraber et al., 2003; Turner et al., 2003; Chen et al., 2005). Appropriate estimates of TBW pool size should be used for larger animals such as swine and cows. Changes in body composition, for example, increasing fat mass with age in swine (Shields et al., 1983) or the extreme obesity seen in ob/ob mice will necessitate appropriate adjustments in estimates of TBW pool size. It has been shown that equilibration with TBW occurs in rats within 10 min of an i.p. injection (Turner et al., 2003).
The body water enrichment using this protocol in rats is stable over several months as shown in Figure 1a (Neese et al., 2002). Similar stability has been obtained in humans (Strawford et al., 2004). A die-away curve of body 2H2O enrichments after cessation of 4% 2H2O intake in mice (Figure 1b) shows that by 2 wk the body water enrichment has gone to essentially zero. In mice, the T1/2 of body water has been estimated to be approximately 2.5 d (Brunengraber et al., 2003).
View larger version (11K):
[in this window]
[in a new window]
|
Figure 1. Total body water 2H2O enrichment in rodents. A) The time course of 2H2O enrichment in body water of rats maintained on 4% 2H2O in drinking water; B) The delabeling, or die-away, curve of body 2H2O enrichment after cessation of 4% 2H2O intake in mice. From Neese et al. (2002).
|
|
In summary, 2H2O is an easy-to-administer tracer that results in a stable precursor pool for the weeks or months necessary to study adipose tissue kinetics. Although the relationship between administered water enrichment and final body water enrichment will need to be determined, available data on body composition and water turnover should allow for easy adaptation of the techniques discussed here to larger animals.
|
METHODS FOR THE MEASUREMENT OF DNL
|
|---|
The use of stable isotopes to study DNL dates back 70 yr to the classic work of Schoenheimer and Rittenberg (1936). However, it was not until much more recently that its application became practical for truly quantitative animal and human studies in vivo. One of the difficulties of stable isotope studies is the accurate determination of the true precursor pool enrichment when that precursor is biochemically unavailable (e.g., cellular acetyl coenzyme A for fatty acid synthesis). Mass isotopomer distribution analysis (MIDA), introduced in 1992 (Hellerstein and Neese, 1992; Kelleher and Masterson, 1992; Lee et al., 1992) and described in more detail elsewhere in these symposium proceedings, is a method of determining true precursor enrichment based upon the mathematics of combinatorial probabilities. Briefly, determinations of true precursor enrichment are made based on the following. Polymers are composed of repeating monomeric subunits. In the case of a 16-carbon palmitate with 13C-labeled acetate as a tracer, the number of repeating subunits, n, is 8. As an example, Figure 2 shows the formation of molecular species containing between 0 and 4 labeled subunits. The ratio of unlabeled (M0), singly labeled (M1), doubly labeled (M2) etc. species to one another (i.e., the isotope pattern in the polymer) differs between the natural abundance and the labeled products. This isotope pattern in an aggregate of subunits is uniquely determined by the proportion of labeled monomers in the precursor pool and n, and follows the principles of the binomial or multinomial expansion. Therefore, by comparing the measured isotope pattern in the polymer, as assessed by mass spectrometry, with expected statistical distributions for a given n (8 in this example), one can determine the isotopic enrichment of the precursor pool (10% in this case). Although several methods have been presented to do the actual calculations, a comparison of these methods in very low density lipoprotein (VLDL) palmitate revealed that, when correctly done, all methods give similar results (Chinkes et al., 1996). A thorough review of MIDA and its applications can be found elsewhere (Hellerstein and Neese, 1999).
View larger version (21K):
[in this window]
[in a new window]
|
Figure 2. The principle of the mass isotopomer distribution analysis technique for measuring biosynthesis of polymers such as fatty acids. Two simulated precursor pools are shown for A) natural abundance, and B) 10% enrichment. The open box represents the natural abundance enriched subunits. The hatched box represents the excess enrichment from subunits enriched from the administered tracer. M0 = unlabeled, M1 = singly labeled, M2 = doubly labeled. The numbers shown are for illustrative purposes only. From Hellerstein and Neese (1999).
|
|
The use of 13C-labeled acetate as a tracer has provided much insight into fatty acid metabolism. However, its expense and inefficient equilibration with the precursor pool in adipose tissue precludes its use for studies of TG metabolism in adipose tissue itself. In addition, zonation across the liver may result in gradients of precursor enrichment (Bederman et al., 2004). An alternative tracer is 2H2O. However, the determination of true precursor enrichment becomes more complex with hydrogen labeling. The palmitate molecule has 31 potentially labeled hydrogens and hence, a maximum n of 31 (Figure 3; the hydroxyl hydrogen is lost during derivatization). However, not all of these hydrogens are equivalent. One hydrogen on each of the even-numbered carbons is contributed directly from water (n = 7). All of the hydrogens on the odd-numbered carbons arise from NADPH (n = 14) (Foster and Katz, 1966; Jungas, 1968). The remaining hydrogens, the methyl hydrogens, and the other hydrogens on the even carbons, come from acetyl-coA (n = 10). There is not complete equilibration between TBW 2H enrichment and 2H enrichment in these latter 2 pools. Acetyl-coA enrichment will depend on its source (Katz and Rognstad, 1966). Similarly, NADPH derived from the pentose cycle will not be in exchange with water (Katz and Rognstad, 1966); however, NADPH derived from the malic enzyme pathway will be. It has been estimated in adipose that at most, 60% of NADPH is derived from the pentose phosphate cycle (Jungas, 1968). Given the complex sources of fatty acid hydrogens, each with a potentially different precursor enrichment, any quantitatively accurate estimate of DNL using heavy water will have to account for these factors.
View larger version (7K):
[in this window]
[in a new window]
|
Figure 3. The structure of methyl palmitate derivative showing 31 potentially labeled hydrogens. R = methyl group. Bold hydrogens (n = 7) are from water, italicized hydrogens (n = 14) are from NADPH, and the other hydrogens (n = 10) are from acetate.
|
|
The first description of the measurement of DNL in humans using 2H2O was by Hachey in 1989. Fatty acid synthesis in the mammary gland was measured in breast milk after a single oral dose of 2H2O. Body water enrichments of 0.05 to 0.1% were achieved resulting in fatty acid enrichments just at the analytic limit for accurate mass spectrometry quantitation. To determine the theoretical maximum possible enrichment, they assumed 2 discrete populations of hydrogens on the fatty acid – 1 in complete exchange with body water and 1 carbon-bound with no exchange. The authors do not discuss how many hydrogens they assign to each population so it is unclear if their calculations will be biased in one direction or another.
Leitch and Jones (1991) used a different strategy. They orally administered heavy water for 2 d and looked at the enrichment in plasma TG using isotope ratio mass spectrometry, a more sensitive method than standard gas chromatography/mass spectrometry (GC/MS). Although this method provided some of the first data on DNL in humans, actual estimates of the fractional synthesis rate were based on assumptions that have since proven incorrect and that lead to overestimation of DNL. The fractional synthesis rate was calculated as the ratio of the TG enrichment to plasma water enrichment with a correction factor of 0.477. This correction factor was based on a presumed 2H:C ratio of 0.87 (corresponding to a palmitate n of 14) based on work by Jungas using tritiated water (Jungas, 1968). However, Jungas showed there was a significant isotope effect and that a correction factor of 1.28 was more appropriate for deuterium labeling (n = 20.5). Secondly, the synthesis rates of specific fatty acids were not analyzed experimentally and the correction is based on a theoretical TG with 3 17-carbon fatty acids. In addition, enrichment in the glycerol moiety is assumed to be in the same ratio as the fatty acid portion of the TG (i.e., 2H:C ratio of 0.87). However, the older literature suggests a ratio of 1.1 (Jungas, 1968). More recent data show that the enrichment in this portion of the molecule can be highly variable (Chen et al., 2005). Finally, it is assumed that the 2H:C ratio is constant, which assumes the contribution of the pentose cycle to the NADPH pool is steady when in fact, this may vary among tissues and under different physiologic conditions. Therefore, although studies using this method have provided valuable insight into the role of DNL in human adipose (Jones, 1996), these represent, at best, a qualitative estimate of DNL.
The use of MIDA to assess the precursor enrichment in palmitate allows for the determination of n that is based on the composite contributions of the different hydrogen pools (Lee et al., 1994b). This method has also been used to determine the synthesis rates of other long-chain fatty acids that are typically synthesized via chain elongation (Ajie et al., 1995). Studies that experimentally determine n have shown that n is dependent on conditions and the tissue being examined. Estimates of palmitate n in the literature range from n = 17 in vitro in hepatoma cells (Lee et al., 1994b) to n = 22 in vivo in rat liver (Lee et al., 1994a). It has been hypothesized that in vivo, deuterium labeling of glucose, lactate, and pyruvate can lead to acetate enrichment, increasing n more so than seen in in vitro experiments, and accounting for n of greater than 21. Palmitate n in plasma TG in humans was shown to be 21 (Lee et al., 1994a; Diraison et al., 1996). The n for adipose tissue with 2H2O has not yet been reported. Given the potential variability of n, this value needs to be determined experimentally when new tissues or experimental conditions are being studied to achieve accurate quantitative determinations of DNL.
|
SIGNIFICANCE OF DNL IN HUMANS AND ANIMALS
|
|---|
The most significant finding regarding DNL in humans is that quantitatively there really isn’t much that occurs under usual dietary conditions. It had been long postulated that excess dietary carbohydrate intake would lead to increased DNL and adipose accumulation. In fact, this is not the case. Schwarz et al. (1995), using [1–13C]-acetate, studied the response of hepatic DNL to caloric excess in the form of carbohydrate, as measured in VLDL-TG fatty acids. Although the percentage contribution of DNL to plasma TG-VLDL increased with increased carbohydrate calories (Figure 4), DNL in absolute terms was quantitatively inconsequential—approximately 3.3 g/d (Schwarz et al., 1995). Studies using heavy water done with non-MIDA methods, which potentially overestimate synthesis rates as discussed above, also show that lipogenesis is quantitatively insignificant in humans (Leitch and Jones, 1993; Jones et al., 1995). Factors thought to increase DNL, such as acute alcohol ingestion or decreased meal frequency, did not result in a significant increase in absolute fatty acid synthesis (Jones et al., 1995; Siler et al., 1998).
View larger version (17K):
[in this window]
[in a new window]
|
Figure 4. Effects of carbohydrate intake on de novo lipogenesis (DNL). Fractional, hepatic DNL in fasted and fed states in response to short-term alterations in dietary energy content as measured in palmitate isolated from circulating very low density lipoprotein fatty acids. Percentages refer to an increase or a decrease in calories. CHO = carbohydrate. a–eBars not sharing a common superscript are significantly different (P < 0.05). From Schwarz et al. (1995).
|
|
These studies in humans all focused on hepatic DNL. However, one could postulate that the majority of DNL, in fact, may occur in adipose tissue. Given the much larger pool size of adipose TG compared with hepatic TG, it should not be surprising that it has been difficult to measure adipose DNL easily with acetate due to dilution of tracer and the associated expense. The use of deuterated water has overcome this problem. In an elegant experiment designed to assess the relative contributions of hepatic vs. adipose DNL to adipose fatty acids, Diraison and collegues (2003) gave subjects both [1,2-13C2]acetate, which is efficiently taken up in the liver but not adipose, and 2H2O, which will equilibrate rapidly with the entire body water pool. By subtracting liver DNL from measured adipose DNL they estimated true adipose DNL, which (as with hepatic DNL) was quantitatively insignificant in humans, even under conditions of carbohydrate excess (Diraison et al., 2003). Another study (Strawford et al., 2004) investigated DNL, from any source–liver or adipose, in adipose tissue in humans using 2H2O. The advantage of this study was that the labeling was over the course of many weeks under normal living conditions, which should allow for the evaluation of DNL under usual ad libitum dietary conditions. Although absolute values for DNL were not presented, one can estimate that at most, whole body DNL accounted for 1 g/d of fatty acid synthesis.
In contrast with humans, DNL appears to be very active and quantitatively significant in other animals. In rodents, DNL in adipose and liver can account for over 50% of fatty acids (Lee et al., 1994a; Brunengraber et al., 2003). In the pig, as much as 80% of adipose fatty acids arise from DNL (O’Hea and Leveille, 1969b). Again in contrast with humans, virtually all DNL in both pigs and cows takes place in the adipose tissue (Ballard et al., 1969; O’Hea and Leveille, 1969b). In other animals, such as chickens, the liver is the primary site of DNL (O’Hea and Leveille, 1969a). There is also species variation with respect to the preferred substrate for DNL. In ruminants, the substrate for synthesis of fatty acids in both the liver and adipose is clearly acetate as opposed to glucose (Ballard et al., 1969). This is in contrast to rodents and pigs where glucose is readily and extensively used for fatty acid synthesis in adipose (e.g., Dunshea et al., 1998).
In summary, several tracers have been used for the in vivo measurement of lipogenesis. Acetate has the advantage of perhaps the easiest calculation of DNL rates because n is known. Although acetate has most often been given intravenously, which limits the duration of administration, it can also be given orally allowing for longer studies. However, cost makes long-term administration in large animals prohibitive. Finally, as discussed, not all tissues and species use acetate as the substrate for DNL. This can be an advantage or disadvantage depending on the aim of the experiment. Glucose has the same constraints as acetate with respect to expense and species preference as DNL substrate. In addition, there is a significant loss of label before its use in DNL, resulting in increased needs for administration and cost. The use of 2H2O as a tracer overcomes potential problems with the choice of the wrong tracer (i.e., glucose vs. acetate) for measuring total lipogenesis in a tissue. Deuterated water will measure synthesis from either source. Its relatively low cost and oral administration allow for inexpensive long-term administration. The disadvantage is the need to determine n for new experimental conditions for the most accurate quantitative results.
|
FACTORS AFFECTING DNL
|
|---|
Diverse dietary factors have been shown to affect DNL in humans. Caloric restriction, even to a modest extent, reduces DNL, whereas an increase in fractional DNL is seen with an increase in caloric intake, especially if the excess calories are from carbohydrates (Schwarz et al., 1995; Aarsland et al., 1997; Diraison et al., 2003). This effect is more pronounced if the dietary carbohydrates are in the form of monosaccharides instead of complex carbohydrates (Hudgins et al., 1998). An extremely low fat (5 to 10% of calories), high-carbohydrate diet can increase fractional DNL even under eucaloric conditions (Hachey et al., 1989; Hudgins et al., 1996). Fructose also appears to increase fractional DNL (Faeh et al., 2005). This effect is more pronounced with acute than with chronic ingestion and can be blunted with intake of fish oil (Faeh et al., 2005). Alcohol also increases fractional DNL (Siler et al., 1998). It should be noted, however, that changes in fractional DNL did not translate into significant changes in absolute DNL. Rather these changes are significant because they signal the state of hepatic fuel metabolism. The importance of relative hepatic DNL to diverse physiologic processes has been shown in mouse models with impaired DNL (Chakravarthy et al., 2005). Changes in the relative contribution of DNL to fatty acids reflect the state of TG assembly, TG production, and liver fat metabolism and thus provide useful information.
Nondietary factors have also been shown to affect fractional DNL in both animals and humans. Inflammation and infection have been shown to upregulate DNL via the action of cytokines such as tumor necrosis factor-, IL-1, and IL-6 (Khovidhunkit et al., 2004). Growth hormone has been shown to decrease DNL in multiple animal species (Etherton, 2000) including pigs (Dunshea et al., 1992; Etherton, 2000), goats (Skarda, 1999), and cows (Lanna et al., 1995). However ractopamine, another agent used to increase protein deposition and decrease fat accumulation, did not suppress lipogenesis in pigs (Dunshea et al., 1998). More complete reviews of the regulation of DNL can be found elsewhere (Hellerstein, 1999; Kersten, 2001; Parks, 2002).
|
STABLE ISOTOPE STUDIES USING LABELED FATTY ACIDS
|
|---|
De novo lipogenesis represents only one aspect of fatty acid metabolism, albeit an important one. Stable isotope-labeled fatty acids have been used to look at fatty acid turnover in TG and to assess fatty acid oxidation. Emken and colleagues (1976) first proposed the use of multiply labeled fatty acids to simultaneously study the relative rates of incorporation and removal of different fatty acids into phospholipids and neutral lipids. They studied differences between the cis and trans configuration of 18:1 fatty acids using d2-elaidate and d4-oleate. Similar methods have subsequently been applied to the study of fatty acid metabolism in breast milk (Emken et al., 1989), polyunsaturated fatty acid metabolism (Emken, 2001), and fatty acid position specificity in TG (Emken et al., 2004). Fatty acid oxidation has been measured by the administration of 13C-labeled fatty acids followed by measurement of CO2 in breath (e.g., Jones et al., 1985; Demmelmair et al., 1997). More recently, the administration of 2H-labeled fatty acids followed by the measurement of 2H2O in urine has been proposed for measurement of fatty acid oxidation (Votruba et al., 2001). 2H-Labeling overcomes many of the limitations of 13C-labeling—difficulties with breath collection, need for a metabolic cart, need for frequent sampling, and confounding isotopic exchange in the TCA cycle. The use of labeled fatty acids as described in these studies allows for the investigations of multiple aspects of fatty acid metabolism.
|
METHODS FOR THE MEASUREMENT OF TG SYNTHESIS
|
|---|
Although much has been learned about adipose kinetics from studies looking at DNL, this measurement has certain limitations. Fatty acid synthesis is not synonymous with the turnover of the intact triglyceride molecule. The fatty acids from TG can arise from dietary fatty acids or via DNL, either in the adipose or liver, and the contribution to different TG pools from each of these sources can vary depending on diet and other factors. Furthermore, partial lipolysis and reesterification further complicate the contribution of newly synthesized fatty acids to TG synthesis. Finally, measurement of DNL is often done for palmitate alone. Because the contribution of palmitate to TG varies (anywhere from 19 to 30% depending on species and dietary factors; Gurr, 1992), so to will estimates of TG synthesis based on palmitate synthesis.
Until recently, estimates of TG synthesis per se were based on labeling of the glycerol moiety of TG using [13C]- [14C]-glucose (e.g., Dunshea et al., 1992 in pigs). However, given the inefficiency of glucose as a precursor and the expense of achieving adequate enrichment in the large, slowly turning over TG pools of adipose tissue, these studies typically have been limited to studying synthesis over the course of several hours. In addition, use of glucose as a precursor is limited by the inability to accurately estimate the true precursor (-GP) enrichment. Labeled glycerol itself is not a viable tracer due to the lack of significant glycerol-kinase activity in adipose. Addressing these limitations are two new methods for the in vivo measurement of TG synthesis in adipose, both involving 2H2O administration (Brunengraber et al., 2003; Turner et al., 2003). These approaches have provided new insights into adipose TG kinetics. The method of Turner et al. (2003) overcomes the problem of true precursor enrichment by experimentally determining n for TG-glycerol in a given tissue using MIDA. Knowledge of the precursor (body water) enrichment and ratio of the excess M2 (EM2) to excess M1 (EM1) in glycerol allows calculation of the actual value of n present under the experimental conditions of the particular study.
There are several potential sources of -GP, the direct intracellular precursor for TG synthesis, as shown in Figure 5. Each of these sources has a different n, and it is impossible to know a priori the quantitative contributions from these sources and, hence, the composite n in any given experiment. However, once n has been determined experimentally, the precursor pool enrichment (body water enrichment) can then be used to calculate the asymptotic value (A
1) for EM1 (Turner et al., 2003). Fractional synthesis can then be calculated based upon the standard precursor product relationship (Wolfe, 1992) to determine new TG synthesized using the equation
View larger version (18K):
[in this window]
[in a new window]
|
Figure 5. Labeling pathways of 2H incorporation into C-H bonds of the glycerol moiety of triglyceride (TG). The pathway of labeled hydrogen incorporation from 2H2O from either glucose (boldface H; TG hydrogens 1, 2, 3, and half of 4), glyceroneogenesis (italicized H, TG hydrogens 1, 2, 3, 4, and 5), or glycerol (TG hydrogens 1, 2, 3). Hydrogen exchange between water and C-H bonds in -glycerol-3-phosphate does not occur during the subsequent synthesis of TG. From Chen et al. (2005).
|
|
 |
This method was used to investigate the kinetics of adipose TG in vivo in different adipose depots. It had long been postulated that mesenteric adipose was more metabolically active than other adipose depots; however, this had been hard to prove. A comparison of TG kinetics in the epididymal, retroperitoneal, and mesenteric depots in rats given 4% 2H2O in drinking water (Figure 6) clearly demonstrated that mesenteric TG does in fact turn over more rapidly, with a half-life of only 3.3 d, which was almost 5 times shorter than the half-life of epididymal TG (Turner et al., 2003).
View larger version (14K):
[in this window]
[in a new window]
|
Figure 6. Acylglyceride turnover in adipose tissue. The time course of label incorporation into the glycerol moiety of adipose acylglycerides during 2H2O administration in adult rats (n = 4/time point). Different adipose depots are shown: A) epididymal; B) inguinal; and C) mesenteric. ks = fractional synthesis rate; t1/2 = half-life. From Turner et al. (2003).
|
|
The method of Brunengraber et al. (2003) focuses solely on the enrichment of hydrogens on the carbon-1 (C1) of TG-glycerol. As shown in Figure 5
, this carbon will be labeled independently of the pathway to -GP formation. Using this method the authors showed that increased caloric intake on a high-fat (45%) diet in mice resulted in increased absolute TG synthesis in epididymal fat. They also showed that TG synthesis and degradation in growing mice is a very active process with degradation (i.e., lipolysis) occurring at approximately 50% of the rate of synthesis (Brunengraber et al., 2003). It has long been known that there can be very high rates of TG reesterification in adipose (Ball, 1965), and it has been hypothesized that this rapid reesterification allows for acute shifts in energy demands. Evidence for this is shown in lactating goats in which the increased availability of fatty acids in the plasma for milk production was not a result of increased lipolysis but rather was due to decreased reesterification (Dunshea et al., 1990).
The use of long-term labeling with 2H2O has allowed, for the first time, isotopic studies estimating adipose TG turnover in humans (Strawford et al., 2004). Using a 9-wk labeling protocol that achieved body water enrichments of approximately 2%, it was found that TG-glycerol fractional synthesis was 12% after 5 wk and 20% after 9 wk giving a half-life on the order of 200 to 270 d. This correlates well with previous estimates based on changes in adipose fatty acid composition after a change in dietary fatty acids (Hirsch, 1965). Net lipolysis (TG turnover) was 50 to 60 g/d. In an attempt to estimate the contribution of DNL to new TG, rather than total TG, as is typically done in DNL studies, fractional DNL was divided by fractional TG synthesis. This revealed that after 9 wk, DNL contributed approximately 20% of the fatty acids to newly synthesized TG. Further studies investigating this relationship could reveal new insights into the relationship between DNL and actual TG synthesis.
Each method described for the measurement of TG synthesis has advantages and disadvantages. Analysis of TG synthesis based on C1 hydrogens requires a complex sample preparation for GC/MS analysis and loses the information obtained from knowing the source of -GP. In addition, with only 2 hydrogens potentially labeled as opposed to 5, stable isotope requirements will be greater to achieve the isotopic enrichments necessary for analytic accuracy. However, MIDA requires the calculation of n with each change in experimental conditions, which means enrichments must be high enough to allow for analytical accuracy in the measurement of M2. Most importantly, both methods provide a new tool for the measurement of adipose kinetics.
|
METHODS FOR THE MEASUREMENT OF GLYCERONEOGENESIS
|
|---|
Although the measurement of n in the glycerol moiety of TG is necessary for the measurement of TG synthesis using MIDA, this also provides information in its own right, revealing the relative contributions from glyceroneogenesis and glycolysis to the synthesis of -GP. Glyceroneogenesis, defined as the synthesis of -GP from gluconeogenic precursors (i.e., precursors other than glucose or glycerol), was first described almost 40 yr ago (Ballard et al., 1967) and is critical for the extensive recycling of fatty acids through triglycerides seen in adipose. Triglyceride glycerol generated from glyceroneogenesis has an n of 5 (italics in Figure 5), whereas that generated from glucose via glycolysis has an n of 3.5 (bold in Figure 5). By empirically determining n, one can then obtain an estimate of the relative contributions of glyceroneogenesis and glycolysis to -GP.
The relationship between TG glycerol n and various physiologic conditions has been explored (Chen et al., 2005). Using TG glycerol n as a marker of the relative contributions of glycolysis and glyceroneogenesis, it was found that in rodents under normal conditions, liver TG is generated predominantly via glyceroneogenesis (over 65%). In contrast, in the adipose, the contribution from glyceroneogenesis is minimal (17%), and TG-glycerol generation is primarily from glucose. In ruminants, where glucose is thought to have minimal contribution to adipose TG, one might expect to see a much greater contribution from glyceroneogenesis.
These relative fluxes can be experimentally altered. A fructose infusion, flooding liver glycolysis, lowered the relative glyceroneogenic contribution to TG-glycerol n in liver from 66 to 34% (Chen et al., 2005). A low carbohydrate diet, on the other hand, increased adipose tissue glyceroneogenesis almost 3-fold. The effects of rosiglitazone on glyceroneogenesis were also investigated. Rosiglitazone, a PPAR- agonist used as an insulin sensitizer in the treatment of type 2 diabetes mellitus, is known to increase adipose TG deposition and alter adipose triglyceride metabolism (Yki-Jarvinen, 2004). The relative importance of rosiglitazone induced increases in phosphoenolpyruvate kinase, the mediator of the rate-limiting step in glyceroneogenesis (Hanson and Reshef, 2003), vs. glycerol kinase expression and activity had been controversial (Reshef et al., 2003). It was shown that TG glycerol n, and hence, the glyceroneogenic contribution, was significantly increased with rosiglitazone administration. This stimulation of glyceroneogenesis in adipose tissue may therefore be an important factor in the antilipolytic actions of these agents (Chen et al., 2005).
The known differences in regulatory control of phosphoenolpyruvate kinase (Hanson and Reshef, 1997) in adipose vs. liver point to potentially important control points of TG synthesis that merit further investigation. Research on factors affecting the relative contribution of glyceroneogenesis to TG synthesis could provide important insights into whole body and adipose tissue lipid metabolism. Moreover, these data suggest that pharmacological manipulation of glyceroneogenesis may have effects on the adipose distribution as shown with rosiglitazone.
|
SUMMARY AND IMPLICATIONS
|
|---|
Deuterated water labeling provides a unique vehicle for the simultaneous study of multiple aspects of TG metabolism: 1) measurement of DNL; 2) measurement of TG synthesis; 3) determination of the source of -GP for TG synthesis; and 4) determination of percentage DNL in new TG. In addition, although not reviewed here, 2H2O labeling has been widely used for the measurement of cell turnover (Neese et al., 2002) and therefore allows for the simultaneous measurement of adipogenesis (Strawford et al., 2004). These tools provide a new opportunity to study factors affecting and integrating multiple aspects of TG synthesis.
Curiously, the goals with regard to adipose metabolism in animal production are virtually opposite the goals in humans. Whereas avoiding adipose gain in the face of excess calories is the goal in humans, efficient energy use is of prime importance in animals raised for food production. Similarly, although intramyocellular fat is associated with insulin resistance and to be avoided in humans, marbling and intramyocellular fat are associated with improved meat quality in animals. Compared with the direct use of carbohydrates for energy, generation of fat for storage for later use (i.e., DNL) is inherently energetically inefficient with a loss of approximately 10% of caloric content (Ball, 1965). Energy is also lost in the rapid lipolysis, reesterification cycle found in adipose TG (approximately 0.3 kcal/g of TG reesterified; Ball, 1965). Therefore, efforts to decrease the quantitatively significant DNL and TG recycling seen in animals could increase efficiency of energy use. New methods that allow for the in vivo study of new aspects of TG synthesis open the door to new possibilities for understanding and altering fat metabolism in animals and humans.
|
Footnotes
|
|---|
1 Invited review. Presented at the "Stable Isotope Tracer Techniques for Nonruminant Nutrition Research and Their Practical Applications" symposium held at the American Society of Animal Science Annual Meeting, Cincinnati, OH, July 24–28, 2005. 
2 The author thanks Scott Turner, Carine Beysen, and Marc Hellerstein for thoughtful comments on this manuscript, and Coline McConnel for help with manuscript preparation.
3 Corresponding author: emurphy{at}kinemed.com
Received for publication August 19, 2005.
Accepted for publication November 22, 2005.
|
LITERATURE CITED
|
|---|
Aarsland, A., D. Chinkes, and R. R. Wolfe. 1997. Hepatic and whole-body fat synthesis in humans during carbohydrate overfeeding. Am. J. Clin. Nutr. 65:1774–1782.[Abstract/Free Full Text]
Ajie, H. O., M. J. Connor, W. N. Lee, S. Bassilian, E. A. Bergner, and L. O. Byerley. 1995. In vivo study of the biosynthesis of long-chain fatty acids using deuterated water. Am. J. Physiol. 269:E247–E252.[Medline]
Ball, E. G. 1965. Some energy relationships in adipose tissue. Ann. N. Y. Acad. Sci. 131:225–234.[Medline]
Ballard, F. J., R. W. Hanson, and D. S. Kronfeld. 1969. Gluconeogenesis and lipogenesis in tissue from ruminant and nonruminant animals. Fed. Proc. 28:218–231.[Medline]
Ballard, F. J., R. W. Hanson, and G. A. Leveille. 1967. Phosphoenolpyruvate carboxykinase and the synthesis of glyceride-glycerol from pyruvate in adipose tissue. J. Biol. Chem. 242:2746–2750.[Abstract/Free Full Text]
Bauman, D. E. 1976. Intermediary metabolism of adipose tissue. Fed. Proc. 35:2308–2313.[Medline]
Bederman, I. R., A. E. Reszko, T. Kasumov, F. David, D. H. Wasserman, J. K. Kelleher, H. Brunengraber. 2004. Zonation of labeling of lipogenic acetyl-coA across the liver: Implications for studies of lipogenesis by mass isotopomer analysis. J. Biol. Chem. 279:43207–43216.[Abstract/Free Full Text]
Brunengraber, D. Z., B. J. McCabe, T. Kasumov, J. C. Alexander, V. Chandramouli, and S. F. Previs. 2003. Influence of diet on the modeling of adipose tissue triglycerides during growth. Am. J. Physiol. Endocrinol. Metab. 285:E917–E925.[Abstract/Free Full Text]
Chakravarthy, M. V., Z. Pan, Y. Zhu, K. Tordjman, J. G. Schneider, T. Coleman, J. Turk, and C. F. Semenkovich. 2005. "New" hepatic fat activates PPAR-alpha to maintain glucose, lipid, and cholesterol homeostasis. Cell. Metab. 1:309–322.[Medline]
Chen, H. C., and R. V. Farese, Jr. 2005. Inhibition of triglyceride synthesis as a treatment strategy for obesity: Lessons from DGAT1-deficient mice. Arterioscler. Thromb. Vasc. Biol. 25:482–486.[Abstract/Free Full Text]
Chen, J. L., E. Peacock, W. Samady, S. M. Turner, R. A. Neese, M. K. Hellerstein, and E. J. Murphy. 2005. Physiologic and pharmacologic factors influencing glyceroneogenic contribution to tria-cylglyceride glycerol measured by mass isotopomer distribution analysis. J. Biol. Chem. 280:25396–25402.[Abstract/Free Full Text]
Chinkes, D. L., A. Aarsland, J. Rosenblatt, and R. R. Wolfe. 1996. Comparison of mass isotopomer dilution methods used to compute VLDL production in vivo. Am. J. Physiol. 271:E373–E383.[Medline]
Demmelmair, H., T. Sauerwald, B. Koletzko, and T. Richter. 1997. New insights into lipid and fatty acid metabolism via stable isotopes. Eur. J. Pediatr. 156(Suppl. 1):S70–S74.[Medline]
Diraison, F., C. Pachiaudi, and M. Beylot. 1996. In vivo measurement of plasma cholesterol and fatty acid synthesis with deuterated water: Determination of the average number of deuterium atoms incorporated. Metabolism 45:817–821.[Medline]
Diraison, F., V. Yankah, D. Letexier, E. Dusserre, P. Jones, and M. Beylot. 2003. Differences in the regulation of adipose tissue and liver lipogenesis by carbohydrates in humans. J. Lipid Res. 44:846–853.[Abstract/Free Full Text]
Dunshea, F. R., A. W. Bell, and T. E. Trigg. 1990. Non-esterified fatty acid and glycerol kinetics and fatty acid re-esterification in goats during early lactation. Br. J. Nutr. 64:133–145.[Medline]
Dunshea, F. R., D. M. Harris, D. E. Bauman, R. D. Boyd, and A. W. Bell. 1992. Effect of porcine somatotropin on in vivo glucose kinetics and lipogenesis in growing pigs. J. Anim. Sci. 70:141–151.[Abstract]
Dunshea, F. R., B. J. Leury, A. J. Tilbrook, and R. H. King. 1998. Ractopamine increases glucose turnover without affecting lipogenesis in the pig. Aust. J. Agric. Res. 49:1147–1152.
Emken, E. A. 2001. Stable isotope approaches, applications, and issues related to polyunsaturated fatty acid metabolism studies. Lipids 36:965–973.[Medline]
Emken, E. A., R. O. Adlof, S. M. Duval, J. M. Shane, P. M. Walker, and C. Becker. 2004. Effect of triacylglycerol structure on absorption and metabolism of isotope-labeled palmitic and linoleic acids by humans. Lipids 39:1–9.[Medline]
Emken, E. A., R. O. Adlof, D. L. Hachey, C. Garza, M. R. Thomas, and L. Brown-Booth. 1989. Incorporation of deuterium-labeled fatty acids into human milk, plasma, and lipoprotein phospholipids and cholesteryl esters. J. Lipid Res. 30:395–402.[Abstract]
Emken, E. A., W. K. Rohwedder, and H. J. Dutton. 1976. Dual-labeled technique for human lipid metabolism studies using deuterated fatty acid isomers. Lipids 11:135–142.[Medline]
Etherton, T. D. 2000. The biology of somatotropin in adipose tissue growth and nutrient partitioning. J. Nutr. 130:2623–2625.[Abstract/Free Full Text]
Faeh, D., K. Minehira, J. M. Schwarz, R. Periasami, P. Seongsu, and L. Tappy. 2005. Effect of fructose overfeeding and fish oil administration on hepatic de novo lipogenesis and insulin sensitivity in healthy men. Diabetes 54:1907–1913.[Abstract/Free Full Text]
Foster, D. W., and J. Katz. 1966. The distribution of tritium in fatty acids synthesized from tritiated glucose and tritiated water by rat adipose tissue. Biochim. Biophys. Acta 125:422–427.[Medline]
Gurr, M. I. 1992. Types of fat in the body and their functions. Pages 7–20 in Role of fats in food and nutrition. M. I. Gurr, ed. Elsevier Applied Science, London, UK.
Hachey, D. L., G. H. Silber, W. W. Wong, and C. Garza. 1989. Human lactation. I: Endogenous fatty acid synthesis by the mammary gland. Pediatr. Res. 25:63–68.[Medline]
Hanson, R. W., and L. Reshef. 1997. Regulation of phosphoenolpyruvate carboxykinase (GTP) gene expression. Annu. Rev. Biochem. 66:581–611.[Medline]
Hanson, R. W., and L. Reshef. 2003. Glyceroneogenesis revisited. Biochimie 85:1199–1205.[Medline]
Hellerstein, M. K. 1999. De novo lipogenesis in humans: Metabolic and regulatory aspects. Eur. J. Clin. Nutr. 53(Suppl. 1):S53–S65.[Medline]
Hellerstein, M. K., and R. A. Neese. 1992. Mass isotopomer distribution analysis: A technique for measuring biosynthesis and turnover of polymers. Am. J. Physiol. 263:E988–1001.[Medline]
Hellerstein, M. K., and R. A. Neese. 1999. Mass isotopomer distribution analysis at eight years: Theoretical, analytic, and experimental considerations. Am. J. Physiol. 276:E1146–E1170.[Medline]
Hirsch, J. 1965. Fatty acid patterns in human adipose tissue. Pages 181–189 in Handbook of Physiology. G. Cahill and A. E. Renold, ed. Waverley, Baltimore, MD.
Hudgins, L. C., M. Hellerstein, C. Seidman, R. Neese, J. Diakun, and J. Hirsch. 1996. Human fatty acid synthesis is stimulated by a eucaloric low fat, high carbohydrate diet. J. Clin. Invest. 97:2081–2091.[Medline]
Hudgins, L. C., C. E. Seidman, J. Diakun, and J. Hirsch. 1998. Human fatty acid synthesis is reduced after the substitution of dietary starch for sugar. Am. J. Clin. Nutr. 67:631–639.[Abstract]
Jones, P. J. 1996. Tracing lipogenesis in humans using deuterated water. Can. J. Physiol. Pharmacol. 74:755–760.[Medline]
Jones, P. J., G. L. Namchuk, and R. A. Pederson. 1995. Meal frequency influences circulating hormone levels but not lipogenesis rates in humans. Metabolism 44:218–223.[Medline]
Jones, P. J., P. B. Pencharz, and M. T. Clandinin. 1985. Whole body oxidation of dietary fatty acids: Implications for energy utilization. Am. J. Clin. Nutr. 42:769–777.[Abstract/Free Full Text]
Jungas, R. L. 1968. Fatty acid synthesis in adipose tissue incubated in tritiated water. Biochemistry 7:3708–3717.[Medline]
Katz, J., and R. Rognstad. 1966. The metabolism of tritiated glucose by rat adipose tissue. J. Biol. Chem. 241:3600–3610.[Abstract/Free Full Text]
Kelleher, J. K., and T. M. Masterson. 1992. Model equations for condensation biosynthesis using stable isotopes and radioisotopes. Am. J. Physiol. 262:E118–E125.[Medline]
Kersten, S. 2001. Mechanisms of nutritional and hormonal regulation of lipogenesis. EMBO Rep. 2:282–286.[Medline]
Khovidhunkit, W., M. S. Kim, R. A. Memon, J. K. Shigenaga, A. H. Moser, K. R. Feingold, and C. Grunfeld. 2004. Effects of infection and inflammation on lipid and lipoprotein metabolism: Mechanisms and consequences to the host. J. Lipid Res. 45:1169–1196.[Abstract/Free Full Text]
Lanna, D. P., K. L. Houseknecht, D. M. Harris, and D. E. Bauman. 1995. Effect of somatotropin treatment on lipogenesis, lipolysis, and related cellular mechanisms in adipose tissue of lactating cows. J. Dairy Sci. 78:1703–1712.[Abstract]
Lee, W. N., S. Bassilian, H. O. Ajie, D. A. Schoeller, J. Edmond, E. A. Bergner, and L. O. Byerley. 1994a. In vivo measurement of fatty acids and cholesterol synthesis using D20 and mass isotopomer analysis. Am. J. Physiol. 266:E699–E708.[Medline]
Lee, W. N., S. Bassilian, Z. Guo, D. Schoeller, J. Edmond, E. A. Bergner, and L. O. Byerley. 1994b. Measurement of fractional lipid synthesis using deuterated water (2H2O) and mass isotopomer analysis. Am. J. Physiol. 266:E372–E383.[Medline]
Lee, W. N., E. A. Bergner, and Z. K. Guo. 1992. Mass isotopomer pattern and precursor-product relationship. Biol. Mass Spectrom. 21:114–122.[Medline]
Leitch, C. A., and P. J. Jones. 1993. Measurement of human lipogenesis using deuterium incorporation. J. Lipid Res. 34:157–163.[Abstract]
Marshall, D. M. 1994. Breed differences and genetic parameters for body composition traits in beef cattle. J. Anim. Sci. 72:2745–2755.[Abstract]
Neese, R. A., L. M. Misell, S. Turner, A. Chu, J. Kim, D. Cesar, R. Hoh, F. Antelo, A. Strawford, J. M. McCune, M. Christiansen, and M. K. Hellerstein. 2002. Measurement in vivo of proliferation rates of slow turnover cells by 2H2O labeling of the deoxyribose moiety of DNA. Proc. Natl. Acad. Sci. USA 99:15345–15350.[Abstract/Free Full Text]
O’Hea, E. K., and G. A. Leveille. 1969a. Lipid biosynthesis and transport in the domestic chick (Gallus domesticus). Comp. Biochem. Physiol. 30:149–159.[Medline]
O’Hea, E. K., and G. A. Leveille. 1969b. Significance of adipose tissue and liver as sites of fatty acid synthesis in the pig and the efficiency of utilization of various substrates for lipogenesis. J. Nutr. 99:338–344.[Abstract/Free Full Text]
Parks, E. J. 2002. Dietary carbohydrate’s effects on lipogenesis and the relationship of lipogenesis to blood insulin and glucose concentrations. Br. J. Nutr. 87(Suppl. 2):S247–S253.[Medline]
Popkin, B. M. 2004. The nutrition transition: An overview of world patterns of change. Nutr. Rev. 62:S140–S143.[Medline]
Reshef, L., Y. Olswang, H. Cassuto, B. Blum, C. M. Croniger, S. C. Kalhan, S. M. Tilghman, and R. W. Hanson. 2003. Glyceroneogenesis and the triglyceride/fatty acid cycle. J. Biol. Chem. 278:30413–30416.[Free Full Text]
Schoenheimer, R., and D. Rittenberg. 1936. Deuterium as an indicator in the study of intermediary metabolism via synthesis and destruction of fatty acids in the organism. J. Biol. Chem. 114:381–396.[Free Full Text]
Schwarz, J. M., R. A. Neese, S. Turner, D. Dare, and M. K. Hellerstein. 1995. Short-term alterations in carbohydrate energy intake in humans. Striking effects on hepatic glucose production, de novo lipogenesis, lipolysis, and whole-body fuel selection. J. Clin. Invest. 96:2735–2743.[Medline]
Shields, R. G., Jr., D. C. Mahan, and P. L. Graham. 1983. Changes in swine body composition from birth to 145 kg. J. Anim. Sci. 57:43–54.[Abstract/Free Full Text]
Siler, S. Q., R. A. Neese, E. J. Parks, and M. K. Hellerstein. 1998. VLDL-triglyceride production after alcohol ingestion, studied using [2-13C1] glycerol. J. Lipid Res. 39:2319–2328.[Abstract/Free Full Text]
Top
Abstract
INTRODUCTION
