Abstract
Labelling molecules with stable isotopes to create tracers has become a gold-standard method to study the metabolism of lipids and lipoproteins in humans. There are a range of techniques which use stable isotopes to measure fatty acid flux and oxidation, hepatic fatty synthesis, cholesterol absorption and synthesis and lipoprotein metabolism in humans. Stable isotope tracers are safe to use, enabling repeated studies to be undertaken and allowing studies to be undertaken in children and pregnant women. This review provides details of the most appropriate tracers to use, the techniques which have been developed and validated for measuring different aspects of lipid metabolism and some of the limitations of the methodology.
Introduction
The global increase in obesity has led to an increased focus on understanding the regulation of lipid metabolism. Obesity is associated with insulin resistance, hyperlipidaemia, non-alcoholic fatty liver disease (NAFLD), type 2 diabetes mellitus and atherosclerotic cardiovascular disease (Hubert et al. 1983, de Marco et al. 1999). Fatty acids are one of the mediators of insulin resistance (Boden 1998), and an imbalance in fatty acid metabolism results in the accumulation of ectopic fat, which can lead to NAFLD, fatty infiltration of the pancreas and epicardial fat (Morelli et al. 2013). Hyperlipidaemia, i.e. increased levels of triacylglycerol (TAG) and cholesterol, is a major risk factor for atherosclerosis, the major cause of cardiovascular disease (Duffield et al. 1983). This review will provide an overview of the techniques that can be used to measure lipid metabolism using stable isotopes in humans.
The concentration of lipids in a blood sample can provide information about whether levels are abnormal, but it cannot provide information about the underlying mechanisms for any abnormality. The concentration of any lipid in a blood sample is a product of its local synthesis and influx into the circulation, absorption from dietary sources and efflux (and local degradation) over time. Therefore, a high concentration of plasma cholesterol, for example, could be a result of its increased synthesis and/or absorption or decreased removal from the circulation. To gain an understanding of lipid metabolism and how it is regulated, we need to be able to measure these fluxes. To do this, the lipid of interest needs to be labelled so that it is indistinguishable from the naturally occurring lipid in the body but is distinct in some way that enables it to be detected and measured accurately. Since the labelled lipid needs to behave metabolically like the unlabelled lipid, the labelling must not change its metabolic characteristics. One way to achieve this is to use stable or radioactive isotopes. Stable isotopes are not a source of ionising radiation, unlike radioactive isotopes. Labelling with stable isotopes has become a gold-standard method to study the metabolism of lipids and lipoproteins in humans. The quantities of stable isotopes that are administered are non-toxic, and studies can be safely undertaken in humans, including pregnant women and children.
Theory
Some elements are composed of atoms that are chemically identical but have a different mass due to different numbers of neutrons. These are termed isotopes. In the biological environment, the most abundant isotopes of the major elements, i.e. hydrogen (H), carbon (C), nitrogen (N) and oxygen (O), are 1H, 12C, 14N and 16O. These are stable, but there are also other stable isotopes of these elements which are much less common (ranging from 0.02 to 1.1%). These are 2H (deuterium), 13C, 15N, 17O and 18O. In this review the term ‘stable isotope’ will refer to these less common isotopes. When one of these less abundant isotopes is substituted for the common form in a molecule, this is known as ‘labelling’ and creates a ‘tracer’. The original molecule is known as the tracee. The tracer and tracee have the same chemical properties. The amount of tracer relative to the amount of tracee is expressed as the tracer:tracee ratio (TTR). TTR corrected for the background TTR is referred to as enrichment. An alternative way of expressing enrichment is as mole percent excess or atom percent excess (APE). This is the amount of tracer (in moles or atoms) relative to the sum of tracer+tracee multiplied by 100. The different mass of stable isotopes enables them to be detected and accurately measured by mass spectrometry (MS).
One of the first studies where stable isotopes were used to study lipid metabolism was published in 1937 (Schoenheimer 1937). In this study, deuterium-labelled linseed oil was fed to mice, and it was shown that 33% of the labelled fatty acids could be detected in adipose tissue (Schoenheimer 1937). However stable isotope-labelled tracers were not easily obtainable and the mass spectrometers needed for their measurement were costly, so this methodology was of limited use for most researchers. However there was a plentiful supply of radioisotopes, which could be used for creating radioactive lipid tracers. These could be measured by scintillation counters which were readily available. The health risk associated with radioactive tracers in humans and the development of affordable mass spectrometers led to the greater use of stable isotope tracers in the late 1960s and early 1970s. Stable isotopes are now widely used for lipid tracer studies, although many of the techniques that are used were developed using radioactive tracers.
Measurement
Isotopic enrichment of biological samples can be measured by nuclear magnetic resonance or MS; the latter is often interfaced with an instrument which first separates the molecule of interest, e.g. gas chromatography (GC), liquid chromatography etc. In GCMS and LCMS the labelled molecule, e.g. [1-13C]palmitate (the tracer), is measured relative to the unlabelled palmitate (i.e. the tracee). A major advantage of stable isotopes over radioactive isotopes is that both tracer and tracee are measured simultaneously with very high precision. An isotope ratio mass spectrometer (IRMS) measures the isotopic enrichment of a gas. This technique is very sensitive and so can be used when enrichment is low. Samples are converted to gases through combustion or pyrolysis, which are then separated. Samples containing 13C-labelled molecules can be converted to CO2 by combustion, while deuterium-containing molecules can be converted to hydrogen by pyrolysis. An IRMS is also used for measuring expired 13CO2 generated from the oxidation of a 13C-labelled molecule and for measuring the enrichment of labelled plasma water (2H2O) which is used in the measurement of fatty acid synthesis. For more details on measurement techniques see Schierbeek et al. (2012).
Lipid metabolism can be measured by the administration of a tracer either orally or intravenously. There is a natural abundance of stable isotopes, so any isotopic tracer administered will add to the existing background. It is therefore essential that the background levels are measured before administering the tracer. Because stable isotope tracers have mass, the quantities used should be low so that the pool size of the tracee is not affected, as this would perturb the kinetics of the system being studied. This requirement is usually fulfilled, since measurement techniques are very sensitive.
Tracer supply and choice
Stable isotope tracers can be purchased from a number of suppliers. For human studies they are relatively expensive because of the quantity required. The cost varies depending on the number of labels and the site of labelling in the molecule. If they are to be administered intravenously in humans, they must be either purchased sterilised and prepared for human use or prepared in this way in a hospital pharmacy. This adds another layer of cost.
The standard notation for a stable isotope tracer indicates the mass number as a superscript to the left of the atom and the number of the atoms substituted with the isotope as a subscript to the right; for example, [2H4]palmitic acid has four atoms of 2H. Uniformly labelled, represented by ‘U’, means all the atoms are substituted with the isotope. Thus [U-13C]palmitic acid indicates that every carbon has been replaced by 13C. The choice of tracer is determined by the level of sensitivity required, whether oxidation rate is to be measured, whether other tracers are to be administered in the same study, whether tracer recycling is likely to occur and other factors such as cost. More details on the most appropriate tracers to use are given when describing each method below.
Stable isotope techniques for measuring lipid kinetics
There are four techniques which are used for measuring lipid kinetics: i) dynamic isotope dilution; ii) the product precursor method; iii) tracer conversion; and iv) the dual isotope method. The choice of technique is dependent on the metabolic pathway to be measured. Some of the tracers used to study lipid metabolism, and the points where they are utilised, are shown in Fig. 1.
Sites of utilisation of palmitate, glycerol, water and acetate tracers used for the measurement of lipid metabolism.
Citation: Journal of Endocrinology 226, 3; 10.1530/JOE-14-0610
Adipose tissue lipolysis
The technique of dynamic isotope dilution can be used to measure whole body production rates of metabolites secreted into the circulation. Since the whole body production rate of non-esterified fatty acids (NEFA) into the blood from tissues is mainly from peripheral adipose tissue in the fasting state, this is used as an index of whole body adipose tissue lipolysis of stored TAG. Although i.v. lipolysis of plasma lipids can contribute, in a fasting state the former is considered to make only a minor contribution (Wolfe et al. 1985). In a postprandial state this becomes more significant (Evans et al. 1999). In a fasting state the rate of production of fatty acids is equal to the rate of whole body uptake, which can also be of interest.
A palmitate tracer is most commonly used for measuring NEFA production rate because it is the cheapest tracer. It is assumed that palmitate is representative of all fatty acids. The tracer can be administered as an i.v. bolus or as a constant i.v. infusion. The latter is the most frequently used, since a bolus requires very rapid blood sampling, as the clearance of NEFA from the circulation is very fast (Mittendorfer et al. 2003). With the constant infusion method, a known amount of the fatty acid tracer, e.g. [1-13C]palmitate, is infused intravenously at a constant rate. Samples are collected ideally from arterialised venous blood or from an artery. In a metabolic steady state, e.g. during fasting, the TTR or APE increases with time until a plateau is reached indicating that an isotopic steady state is achieved (Fig. 2). At this point the rate of removal of the tracer from the system is equal to the rate that it appears, i.e. the infusion rate. In the plasma pool the tracer is mixed with tracee entering the system from adipose tissue, and measurement of its dilution provides a measure of the rate at which palmitate is appearing in the body pool. This technique requires the unlabelled palmitate to be constant during the study, i.e. in a steady state. It is assumed that the tracer is diluted in a single body pool, e.g. plasma. The dilution of the tracer by the tracee at isotopic steady state is a measure of the rate of appearance (Ra) of palmitate into the pool, which in a steady state is equal to the rate of loss from the pool (also known as the flux) (Fig. 3).
Schematic example of a constant infusion of tracer to achieve an isotopic steady state.
Citation: Journal of Endocrinology 226, 3; 10.1530/JOE-14-0610
Calculating rate of appearance (Ra) at isotopic steady state. Rd, rate of disappearance; black circles represent the tracer; white circles represent the tracee. Reproduced, with permission, from Wolfe RR & Chinkes DL (2004) Isotope Tracers in Metabolic Research: Principles and Practice of Kinetic Analysis, 2nd Edition. Copyright 2004 Wiley-Blackwell.
Citation: Journal of Endocrinology 226, 3; 10.1530/JOE-14-0610
where SS=isotopic steady state
To convert palmitate Ra to NEFA Ra, the concentration of palmitate relative to total NEFA needs to be measured. Other fatty acids can be used as tracers; palmitate, oleate and linoleate have all been shown to provide good estimates of NEFA Ra, whereas myristate and stearate do not (Mittendorfer et al. 2003). Since long-chain fatty acids are insoluble in aqueous solution, the stable isotope tracer of a fatty acid salt, e.g. potassium palmitate, is complexed with human albumin to aid solubility. Deuterated tracers are cheaper, but if [U-13C]palmitate is used, the infusion rate can be reduced if GC-combustion IRMS is used to measure enrichment. This reduces the amount of albumin which needs to be infused. If [U-13C]palmitate is used, palmitate oxidation rate can also be measured.
NEFA production rate can underestimate adipose tissue lipolysis, because some NEFA released from lipolysis is re-esterified within adipose tissue. The production rate of glycerol measured with a glycerol tracer provides a more accurate measure of lipolysis, since glycerol kinase levels are very low in adipose tissue, so there is very little recycling (Miles et al. 2004). As with using NEFA production rate as a measure of lipolysis, i.v. lipolysis of plasma lipids can also contribute some glycerol to the plasma pool, although in a fasting state this would be a minor contribution (Robinson & Newsholme 1967). The technique used is identical to the one described above for palmitate. There are several options in the choice of tracer: [2H5]glycerol, 2-13C1 and U-13C glycerol have all been shown to give similar measurements of whole body glycerol production rate (Wolfe 1992).
The isotope dilution technique can also be used in the non-steady state. With the tracee in a steady state, e.g. during fasting, an isotopic steady state is achieved first with the tracer infusion. The steady state can then be disturbed by, for example, the subject exercising or with the infusion of a hormone. The changes in TTR and concentration with time are used to calculate the change in the flux of the tracee using non-steady-state equations or mathematical modelling (Steele et al. 1965). Using this technique, the insulin sensitivity of lipolysis can be measured by combining a constant i.v. infusion of labelled glycerol or labelled palmitate with a low-dose insulin infusion (Shojaee-Moradie et al. 2007) (Fig. 4).
Glycerol production rate in a healthy subject measured with a constant infusion of [2H5]glycerol during fasting and following an infusion of insulin. The decrease in glycerol production rate is a measure of the insulin sensitivity of lipolysis.
Citation: Journal of Endocrinology 226, 3; 10.1530/JOE-14-0610
NEFA Ra can also be measured during the postprandial period, using non-steady-state equations. However, in the postprandial period, some fatty acids hydrolysed from dietary fat (chylomicrons) can directly enter the plasma pool, a process known as spillover, so NEFA Ra is a measure of both lipolysis and spillover (Bickerton et al. 2007).
The isotope dilution technique can be combined with the measurement of arterio–venous balance measurements in specific tissues, e.g. subcutaneous adipose tissue. This can provide measurements of tissue-specific rates of production and uptake (Bickerton et al. 2007). This method requires the measurement of blood flow through the tissue.
Fatty acid oxidation
Whole body NEFA oxidation can be measured using the technique of tracer conversion. If 13C-labelled palmitate is administered, the appearance of 13C in CO2 can provide a measure of palmitate oxidation rate. Small bags or tubes are used to collect expired air. The 13CO2 content can then be measured by IRMS. To calculate the 13CO2 expiry rate, CO2 production rate (VCO2) must also be measured. This can be obtained using a gas analyser when collecting the breath samples. A quantitative measure of palmitate oxidation rate can be determined if a steady state of 13CO2 is achieved. An isotopic steady state of 13CO2 can take up to 8 h due to a slow rate of exchange of 13CO2 with the bicarbonate pool. This can be shortened by priming the bicarbonate pool by administering an i.v. bolus of 13C sodium bicarbonate when the tracer infusion is initiated. The other consideration with this method is correction for any fixation of 13CO2 in the body bicarbonate pool and loss of 13C during the TCA cycle. A method for correcting for this has been developed using 13C-labelled acetate (Schrauwen et al. 1998). In this method 13C-labelled acetate is infused in a separate study, which mimics the conditions employed for the determination of palmitate oxidation (Schrauwen et al. 1998). Acetate enters the TCA cycle as acetyl CoA, and the difference between the amount of 13C acetate infused and the amount recovered in13CO2 is a measure of the amount of fixation which needs to be corrected for (Fig. 5). The need for an additional study adds an extra burden to this measurement.
Plasma NEFA oxidation calculated using the acetate correction factor when infusing [U-13C]palmitate. The correction factor is the fractional 13C-label recovery in breath CO2, observed after an infusion of [1,2-13C]acetate. Reproduced, with permission, from Schrauwen P, van Aggel-Leijssen DP, van Marken Lichtenbelt WD, van Baak MA, Gijsen AP, Wagenmakers AJ (1998) Validation of the [1,2-13C]acetate recovery factor for correction of [U-13C]palmitate oxidation rates in humans. Journal of Physiology 513 215–223. Copyright 2004, John Wiley and Sons.
Citation: Journal of Endocrinology 226, 3; 10.1530/JOE-14-0610
Lipoprotein metabolism
Lipoprotein particle kinetics can be measured by labelling one of the apolipoprotein components with an isotopically labelled amino acid, e.g. 13C leucine. Lipoprotein TAG kinetics can be measured by labelling TAG in the TAG-rich lipoproteins, i.e. VLDL and/or chylomicrons. To measure the synthesis rates of macromolecules such as proteins and triglycerides, the precursor-product method is used. A stable isotope-labelled precursor is administered either orally or intravenously, and the incorporation of the precursor into the macromolecule is measured over a period of time. If the TTR of the precursor is known or can be measured, the fractional rate of synthesis (FSR) of the product can be determined. In a steady state, i.e. when the protein mass or TG pool is constant, the FSR is equal to the fractional catabolic rate (FCR), a measure of metabolic clearance. FSR and FCR have per unit of time. An absolute secretion rate (ASR) (e.g. mmol/h) is calculated by multiplying the FSR by the product pool size.
Apolipoprotein kinetics
ApoB100 is a large structural protein in VLDL, IDL and LDL. There is one apoB100 molecule/lipoprotein particle. VLDL, IDL and LDL apolipoprotein (apo) B100 FCR and ASR can be measured with either an i.v. bolus or constant infusion of a labelled amino acid (Millar et al. 1995, Demant et al. 1996). Most researchers use [1-13C]leucine or [2H3]leucine, although labelled valine and lysine have also been used.
With the infusion method the rate of increase in apoB enrichment measured relative to the precursor pool is a measure of apoB fractional secretion rate (FSR). With the bolus method, the decrease in apoB enrichment over time is measured. With both methods the change in enrichment with time is not linear, so fitting the data with a linear regression is inappropriate. A monoexponential function can be fitted to the enrichment data (Cummings et al. 1995) to calculate FSR, but a mathematical model can provide a better curve fit and more accurate measurements of FSR (Parhofer et al. 1991). Because VLDL is metabolised to IDL and then LDL in the circulation, and during this process apoB100 is retained within the lipoprotein, if apoB100 enrichment is measured in each fraction, it is possible to describe this whole metabolic pathway using a multicompartmental mathematical model (Fig. 6). A number of different model structures have been proposed (Millar et al. 1995, Duvillard et al. 2000). If VLDL apoB100 enrichment achieves an isotopic steady state during a constant–infusion study, this plateau of enrichment is a measure of the enrichment of the precursor pool. If this is not the case, an alternative measure of the precursor pool is used. Plasma leucine does not provide an appropriate measure of the precursor pool, since hepatic intracellular leucine (the true precursor pool) is diluted by intracellular proteolysis. Within hepatic cells, leucine is deaminated to alpha ketoisocaproate (α KIC), and the enrichment of α KIC, which can be found in the circulation, has been shown to provide a good estimate of hepatic intracellular leucine enrichment (Barazzoni et al. 1999).
Typical curve fit for leucine tracer:tracee ratio in VLDL, IDL and LDL apoB100 using mathematical modelling following an infusion of [1-13C]leucine. VLDL shown as large squares. IDL shown as triangles. LDL shown as circles. The solid lines indicate the curve fit. The modelling provides a measure of VLDL, IDL and LDL apoB100 fractional catabolic rate and production rate.
Citation: Journal of Endocrinology 226, 3; 10.1530/JOE-14-0610
VLDL and IDL apoB100 enrichment can be measured by GCMS, but there is considerable dilution of the tracer in the large pool of LDL apoB100. This necessitates the measurement of LDL apoB100 enrichment by GC-combustion IRMS, which has a much higher sensitivity than GCMS. In several studies VLDL has been subfractionated and the enrichment and kinetics of two forms of VLDL measured: VLDL1, a large triglyceride-rich lipoprotein; and VLDL2, a less triglyceride-rich lipoprotein (Adiels et al. 2005). VLDL2 can be formed by direct secretion from the liver or from catabolism of VLDL1, and both these pathways can be determined by mathematical modelling. Unlike VLDL, LDL has a very slow turnover, so a constant infusion of tracer over a single day provides a limited description of the LDL apoB enrichment curve. An i.v. bolus of tracer provides an alternative approach, which enables LDL apoB enrichment to be measured over several days if volunteers are willing to return to the clinical research centre for a single blood sample over the following 12–14 days. This provides a more accurate measurement of LDL kinetics (Demant et al. 1996).
In studies where constant meal feeding is used to achieve a postprandial steady state, labelling of apoB48, the structural protein of chylomicrons, with a labelled amino acid such as [1-13C]leucine can provide a measure of chylomicron particle kinetics. The methodology is identical to that described for VLDL apoB100. Both apoB48 and apoB100 are isolated by SDS polyacrylamide gel electrophoresis (Lichtenstein et al. 1992).
By labelling the main protein component, apoA1, HDL particle kinetics can be measured (Cohn et al. 1990). This has been measured with a constant i.v. infusion of a labelled amino acid over a single day. However HDL has a half-life of 5 days, which means a very limited description of the apoA1 enrichment curve is gained in 1 day. As with measuring LDL kinetics, the i.v. bolus technique with measurement of enrichment over several days combined with mathematical modelling should provide a more robust measurement of HDL apoA1 kinetics. There are, however, no published studies which have investigated this using stable isotopes.
VLDL and chylomicron TAG kinetics
VLDL TAG kinetics can be measured with an i.v. bolus or constant infusion of labelled glycerol (Sarac et al. 2012) or a fatty acid, usually palmitate (Patterson et al. 2002). Glycerol is the preferred tracer, since recycling of a fatty acid tracer can occur, which can result in an underestimation of VLDL TAG synthesis rate (Patterson et al. 2002). Administration of an i.v. bolus of 2H5 glycerol with measurement of the rise and decline in glycerol enrichment in triglyceride over 8–12 h is the most widely used method (Sarac et al. 2012). The FSR (and FCR) can be determined by mathematical modelling as described above for apoB100 kinetics (Adiels et al. 2005, Parhofer et al. 1991). Depending on the ultracentrifugation methods used to isolate VLDL, either total VLDL TAG kinetics or VLDL1 and VLDL2 TAG kinetics can be measured (Fig. 7).
Typical curve fit of VLDL1 and VLDL2 glycerol tracer:tracee ratio in TAG, following an i.v. bolus of [2H5]glycerol, using mathematical modelling. VLDL1 shown as large squares. VLDL2 shown as small circles. The solid lines indicate the curve fit. The modelling provides a measure of VLDL1 and VLDL2 TAG fractional catabolic rate and production rate. Reproduced, with permission, from Sarac I, Backhouse K, Shojaee-Moradie F, Stolinski M, Robertson MD, Bell JD, Thomas EL, Hovorka R, Wright J & Umpleby AM (2012) Gender differences in VLDL1 and VLDL2 triglyceride kinetics and fatty acid kinetics in obese postmenopausal women and obese men. Journal of Clinical Endocrinology and Metabolism 97 2475–2481. Copyright 2012 The Endocrine Society.
Citation: Journal of Endocrinology 226, 3; 10.1530/JOE-14-0610
An i.v. bolus of 2H5 glycerol has been shown to be incorporated into chylomicron TAG, enabling chylomicron TAG synthesis to be measured in a constant-feeding study (Fig. 8) (Shojaee-Moradie et al. 2013). This is a laborious method because chylomicron particles must be separated from VLDL particles. This cannot be achieved by ultracentrifugation. In the above study an immunoaffinity technique using three antibodies to apoB100 was used to isolate VLDL from chylomicrons (Shojaee-Moradie et al. 2013).
Glycerol tracer:tracee ratio in chylomicron TAG in healthy subjects (white circles) and subjects with MetS (black circles) following an i.v. bolus of [2H5]glycerol (mean±s.e.m.). Mathematical modelling of this data provides a measure of chylomicron TAG fractional catabolic rate and production rate. Reproduced, with permission, from Shojaee-Moradie et al. (2013). Diabetes, American Diabetes Association, 2013. Copyright and all rights reserved. Material from this publication has been used with the permission of American Diabetes Association.
Citation: Journal of Endocrinology 226, 3; 10.1530/JOE-14-0610
Hepatic fatty acid synthesis
The precursor-product method is also used for measuring hepatic fatty acid synthesis. Since fatty acids are synthesised from multiple units of acetyl CoA, an i.v. infusion of 13C acetate (the precursor) should enable the determination of hepatic fatty acid synthesis (frequently referred to as de novo lipogenesis; DNL) if the TTR of the precursor (acetate) is known and the TTR of the product, e.g. palmitate, the major product of hepatic synthesis, is measured in VLDL TAG (which is produced by the liver). However, because there are different pools of acetyl CoA in the liver used for different metabolic pathways, the plasma acetate TTR cannot be assumed to be the true precursor. One way to calculate the enrichment of the true precursor uses mass isotopomer distribution analysis (MIDA). This method determines the precursor enrichment from the pattern of labelling in the product (Hellerstein & Neese 1999). A constant i.v. infusion of the acetate tracer is administered for 6−8 h, with measurement of the enrichment of palmitate in VLDL TAG at the end of the infusion period. The frequency of double-labelled relative to single-labelled palmitate is measured and the TTR of the precursor is calculated using an algorithm. This method has been used in both fasting and feeding studies. In feeding studies repeated mixed meal feeding is used to achieve a postprandial steady state in VLDL TAG (Stanhope et al. 2009). With an infusion period of 8 h, the fractional synthesis of palmitate is determined as
where TTRpalmitate is the enrichment of palmitate in VLDL TAG at time 0 (before the tracer infusion) and 8 h, and TTRPrecursor is the enrichment of the precursor pool (acetate) at 8 h.
An alternative method for measuring hepatic fatty acid synthesis uses labelled water. During palmitate synthesis, H from NADPH and H2O is added, and this can be utilised to measure palmitate synthesis rate (Leitch & Jones 1993). When 2H2O is given orally, it rapidly equilibrates with total body water and 2H is incorporated into palmitate as it is synthesised. A plateau of deuterium enrichment is reached within 12 h. In the most widely used method an oral dose (3 g/kg body weight) of 2H2O which aims to produce 0.45% enrichment of body water is given in the evening, half with the evening meal and half at 22 h. Subjects drink only water (0.45% enriched) to prevent dilution of the labelled body water with ingested water, until after a blood sample is taken the following morning for the measurement of VLDL TAG-palmitate enrichment and plasma water enrichment. There are 31 hydrogens in palmitate which could potentially be labelled. However, not all hydrogens are equivalent. Seven H are directly from H2O, 14 are from NADPH and ten are from acetyl CoA. Complete equilibration with the labelled body water pool does not occur with the latter two pools. It has been shown, using MIDA, that the maximum number of labelled H that can be incorporated into palmitate is 21 (Diraison et al. 1996).
If all VLDL TAG-palmitate are derived from DNL, the enrichment in TAG-palmitate will be equal to the enrichment of the plasma water multiplied by the maximum number of labelled H that can be incorporated.
The maximum theoretical proportion of VLDL TAG-palmitate which can be derived from DNL is 100%. The actual proportion is calculated from the ratio of the observed VLDL TAG-palmitate enrichment (TTR) to the maximum theoretical palmitate enrichment (the theoretical enrichment if all VLDL TAG-palmitate is derived from DNL).
The absolute rate of VLDL TAG synthesis from DNL can be calculated by multiplying %DNL by total VLDL TAG synthesis rate (e.g. measured using a glycerol or pamitate tracer as described in the section on VLDL TAG kinetics).
De novo cholesterol synthesis and absorption from the diet
A high cholesterol level is a strong independent risk factor for cardiovascular disease. Understanding the regulation of cholesterol metabolism is therefore of major importance. Plasma cholesterol is derived from endogenous synthesis and exogenous sources (diet and bile acids). Stable isotope tracers can be used to measure cholesterol absorption from the diet and from de novo synthesis.
Cholesterol absorption
The dual-isotope method is used for measuring oral absorption of cholesterol (Bosner et al. 1993). It is based on a method previously developed by Zilversmit & Hughes (1974) using radioisotopes. An oral dose of 2H5 cholesterol (or 2H6 or 2H7 cholesterol) in a meal and an i.v. injection of 13C5 cholesterol (or 13C6 or 13C7 cholesterol) is administered on day 0, and on day 3 the plasma enrichment of the two tracers is measured. At this point it has been shown that the ratio of enrichment of the oral/i.v. tracer in the plasma is stable (Bosner et al. 1993). A limitation of the method is that the labelled cholesterol absorption is not necessarily the same as cholesterol absorption from food. Percentage cholesterol absorption is calculated from the ratio of plasma 2H5 cholesterol TTR/13C5 cholesterol TTR×100 on day 3. This method has been shown to have good precision and repeatability (Bosner et al. 1993). An alternative method has used 2H5 cholesterol as the tracer and 2H4 sitostanol as a non-absorbable marker given orally for 7 days. Cholesterol absorption was calculated from faecal samples collected over days 4–7 and measured by GCMS (Lutjohann et al. 1993).
Cholesterol synthesis
The precursor-product method is used for measuring free cholesterol synthesis. There are two principal methods which are very similar to methods used for measuring fatty acid synthesis as described above. An i.v. infusion of 13C acetate combined with the MIDA method to determine the true precursor pool has been used to measure the FSR of cholesterol (Neese et al. 1993). To measure the absolute synthesis rate, blood samples can be collected over the next few days to measure the decay of the labelled cholesterol with time. The rate constant for this removal rate provides an estimate of the cholesterol pool size. The absolute synthesis rate of cholesterol can then be calculated as:
The pool size is assumed to be free cholesterol in liver, plasma and red cells, as these free cholesterol pools are in rapid equilibrium in humans. What is measured by this method is the synthesis of cholesterol (which is predominantly in the intestine and liver) which is secreted into the plasma cholesterol pool. Total body cholesterol synthesis is not measured by this method, because tissue cholesterol synthesis that does not leave the cell or equilibrate with the plasma pool is not measured. While some researchers have used an 8-h i.v. infusion of 13C acetate, a 24-h infusion is recommended.
The factor 0.478 is the fraction of hydrogen atoms/cholesterol molecule that can become enriched with deuterium during cholesterol synthesis. This can be converted to an absolute synthesis rate if assumptions are made about the size of the free cholesterol pool.
The FSR measured by an i.v. infusion of 13C acetate and the MIDA method agrees well with FSR measured with oral 2H2O, showing a significant correlation (r=0.84, P=0.0007) in 12 subjects (Di Buono et al. 2000).
Conclusion
Measuring lipid concentrations in blood samples provides only a static measurement of lipid metabolism. This provides inadequate information for understanding lipid metabolism in health and disease in humans. Using stable isotope tracers, the fluxes of molecules through lipid metabolic pathways can be measured in vivo. There are many different methods which can be used, and the choice of method and tracer depends on the lipid metabolite to be studied. The techniques and tracers can be used individually or several techniques and tracers can be combined in a single study enabling multiple metabolic pathways to be measured simultaneously.
Declaration of interest
The author declares that there is no conflict of interest that could be perceived as prejudicing the impartiality of this guideline.
Funding
This work did not receive any specific grant from any funding agency in the public, commercial or not-for-profit sector.
References
Adiels M, Packard C, Caslake MJ, Stewart P, Soro A, Westerbacka J, Wennberg B, Olofsson SO, Taskinen MR & Boren J 2005 A new combined multicompartmental model for apolipoprotein B-100 and triglyceride metabolism in VLDL subfractions. Journal of Lipid Research 46 58–67. (doi:10.1194/jlr.M400108-JLR200)
Barazzoni R, Meek SE, Ekberg K, Wahren J & Nair KS 1999 Arterial KIC as marker of liver and muscle intracellular leucine pools in healthy and type 1 diabetic humans. American Journal of Physiology 277 E238–E244.
Bickerton AS, Roberts R, Fielding BA, Hodson L, Blaak EE, Wagenmakers AJ, Gilbert M, Karpe F & Frayn KN 2007 Preferential uptake of dietary Fatty acids in adipose tissue and muscle in the postprandial period. Diabetes 56 168–176. (doi:10.2337/db06-0822)
Boden G 1998 Free fatty acids (FFA), a link between obesity and insulin resistance. Frontiers in Bioscience 3 169–175.
Bosner MS, Ostlund RE, Osofisan O, Grosklos J, Fritschle C & Lange LG 1993 Assessment of percent cholesterol absorption in humans with stable isotopes. Journal of Lipid Research 34 1047–1053.
Cohn JS, Wagner DA, Cohn SD, Millar JS & Schaefer EJ 1990 Measurement of very low density and low density lipoprotein apolipoprotein (Apo) B-100 and high density lipoprotein Apo A-I production in human subjects using deuterated leucine. Effect of fasting and feeding. Journal of Clinical Investigation 85 804–811. (doi:10.1172/JCI114507)
Cummings MH, Watts GF, Umpleby M, Hennessy TR, Quiney JR & Sönksen PH 1995 Increased hepatic secretion of very-low-density-lipoprotein apolipoprotein B-100 in heterozygous familial hypercholesterolaemia: a stable isotope study. Atherosclerosis 113 79–89. (doi:10.1016/0021-9150(94)05430-Q)
Demant T, Packard CJ, Demmelmair H, Stewart P, Bedynek A, Bedford D, Seidel D & Shepherd J 1996 Sensitive methods to study human apolipoprotein B metabolism using stable isotope-labeled amino acids. American Journal of Physiology 270 E1022–E1036.
Di Buono M, Jones PJ, Beaumier L & Wykes LJ 2000 Comparison of deuterium incorporation and mass isotopomer distribution analysis for measurement of human cholesterol biosynthesis. Journal of Lipid Research 41 1516–1523.
Diraison F, Pachiaudi C & Beylot M 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. (doi:10.1016/S0026-0495(96)90152-3)
Duffield RG, Lewis B, Miller NE, Jamieson CW, Brunt JN & Colchester AC 1983 Treatment of hyperlipidaemia retards progression of symptomatic femoral atherosclerosis. A randomised controlled trial. Lancet 2 639–642. (doi:10.1016/S0140-6736(83)92527-8)
Duvillard L, Pont F, Florentin C, Galland-Jos P, Gambert P & Verges B 2000 Metabolic abnormalities of apolipoprotein B containing lipoproteins in non-insulin dependent diabetes: a stable isotope kinetic study. European Journal of Clinical Investigation 30 685–694.
Evans K, Clark ML & Frayn KN 1999 Effects of an oral and intravenous fat load on adipose tissue and forearm lipid metabolism. American Journal of Physiology 276 E241–E248.
Hellerstein MK & Neese RA 1999 Mass isotopomer distribution analysis at eight years: theoretical, analytic, and experimental considerations. American Journal of Physiology 276 E1146–E1170.
Hubert HB, Feinleib M, McNamara PM & Castelli WP 1983 Obesity as an independent risk factor for cardiovascular disease: a 26-year follow-up of participants in the Framingham Heart Study. Circulation 67 968–977. (doi:10.1161/01.CIR.67.5.968)
Jones PJ, Scanu AM & Schoeller DA 1988 Plasma cholesterol synthesis using deuterated water in humans: effect of short-term food restrictions. Journal of Laboratory and Clinical Medicine 111 627–633.
Leitch CA & Jones PJ 1993 Measurement of human lipogenesis using deuterium incorporation. Journal of Lipid Research 34 157–163.
Lichtenstein AH, Hachey DL, Millar JS, Jenner JL, Booth L, Ordovas J & Schaefer EJ 1992 Measurement of human apolipoprotein B-48 and B-100 kinetics in triglyceride-rich lipoproteins using [5,5,5-2H3]leucine. Journal of Lipid Research 33 907–914.
Lutjohann D, Meese CO, Crouse JR III & von Bergmann K 1993 Evaluation of deuterated cholesterol and deuterated sitostanol for measurement of cholesterol absorption in humans. Journal of Lipid Research 34 1039–1046.
de Marco R, Locatelli F, Zoppini G, Verlato G, Bonora E & Muggeo M 1999 Cause-specific mortality in type 2 diabetes. The Verona Diabetes Study. Diabetes Care 22 756–761. (doi:10.2337/diacare.22.5.756)
Miles JM, Park YS, Walewicz D, Russell-Lopez C, Windsor S, Isley WL, Coppack SW & Harris WS 2004 Systemic and forearm triglyceride metabolism: fate of lipoprotein lipase-generated glycerol and free fatty acids. Diabetes 53 521–527. (doi:10.2337/diabetes.53.3.521)
Millar JS, Lichtenstein AH, Cuchel M, Dolnikowski GG, Hachey DL, Cohn JS & Schaefer EJ 1995 Impact of age on the metabolism of VLDL, IDL, and LDL apolipoprotein B-100 in men. Journal of Lipid Research 36 1155–1167.
Mittendorfer B, Liem O, Patterson BW, Miles JM & Klein S 2003 What does the measurement of whole-body fatty acid rate of appearance in plasma by using a fatty acid tracer really mean? Diabetes 52 1641–1648. (doi:10.2337/diabetes.52.7.1641)
Morelli M, Gaggini M, Daniele G, Marraccini P, Sicari R & Gastaldelli A 2013 Ectopic fat: the true culprit linking obesity and cardiovascular disease? Thrombosis and Haemostasis 110 651–660. (doi:10.1160/TH13-04-0285)
Neese RA, Faix D, Kletke C, Wu K, Wang AC, Shackleton CH & Hellerstein MK 1993 Measurement of endogenous synthesis of plasma cholesterol in rats and humans using MIDA. American Journal of Physiology 264 E136–E147.
Parhofer KG, Hugh P, Barrett R, Bier DM & Schonfeld G 1991 Determination of kinetic parameters of apolipoprotein B metabolism using amino acids labeled with stable isotopes. Journal of Lipid Research 32 1311–1323.
Patterson BW, Mittendorfer B, Elias N, Satyanarayana R & Klein S 2002 Use of stable isotopically labeled tracers to measure very low density lipoprotein-triglyceride turnover. Journal of Lipid Research 43 223–233.
Robinson J & Newsholme EA 1967 Glycerol kinase activities in rat heart and adipose tissue. Biochemical Journal 104 2c–4c.
Sarac I, Backhouse K, Shojaee-Moradie F, Stolinski M, Robertson MD, Bell JD, Thomas EL, Hovorka R, Wright J & Umpleby AM 2012 Gender differences in VLDL1 and VLDL2 triglyceride kinetics and fatty acid kinetics in obese postmenopausal women and obese men. Journal of Clinical Endocrinology and Metabolism 97 2475–2481. (doi:10.1210/jc.2011-3248)
Schierbeek H, van den Akker CH, Fay LB & van Goudoever JB 2012 High-precision mass spectrometric analysis using stable isotopes in studies of children. Mass Spectrometry Reviews 31 312–330. (doi:10.1002/mas.20344)
Schoenheimer R 1937 The investigation of intermediary metabolism with the aid of heavy hydrogen. Bulletin of the New York Academy of Medicine 13 272–295.
Schrauwen P, van Aggel-Leijssen DP, van Marken Lichtenbelt WD, van Baak MA, Gijsen AP & Wagenmakers AJ 1998 Validation of the [1,2-13C]acetate recovery factor for correction of [U-13C]palmitate oxidation rates in humans. Journal of Physiology 513 215–223. (doi:10.1111/j.1469-7793.1998.215by.x)
Shah V, Herath K, Previs SF, Hubbard BK & Roddy TP 2010 Headspace analyses of acetone: a rapid method for measuring the 2H-labeling of water. Analytical Biochemistry 404 235–237. (doi:10.1016/j.ab.2010.05.010)
Shojaee-Moradie F, Baynes KC, Pentecost C, Bell JD, Thomas EL, Jackson NC, Stolinski M, Whyte M, Lovell D & Bowes SB et al. 2007 Exercise training reduces fatty acid availability and improves insulin sensitivity of glucose metabolism. Diabetologia 50 404–413. (doi:10.1007/s00125-006-0498-7)
Shojaee-Moradie F, Ma Y, Lou S, Hovorka R & Umpleby AM 2013 Prandial hypertriglyceridemia in metabolic syndrome is due to an overproduction of both chylomicron and VLDL triacylglycerol. Diabetes 62 4063–4069. (doi:10.2337/db13-0935)
Stanhope KL, Schwarz JM, Keim NL, Griffen SC, Bremer AA, Graham JL, Hatcher B, Cox CL, Dyachenko A & Zhang W et al. 2009 Consuming fructose-sweetened, not glucose-sweetened, beverages increases visceral adiposity and lipids and decreases insulin sensitivity in overweight/obese humans. Journal of Clinical Investigation 119 1322–1334. (doi:10.1172/JCI37385)
Steele R, Bishop JS, Dunn A, Altszuler N, Rathbeb I & Debodo RC 1965 Inhibition by insulin of hepatic glucose production in the normal dog. American Journal of Physiology 208 301–306.
Wolfe RR 1992 Radioactive and stable isotope tracers in biomedicine. In Principles and Practice of Kinetic Analysis. New York, NY, USA: Wiley Liss Inc.
Wolfe RR & Chinkes DL 2004 In Isotope Tracers in Metabolic Research: Principles and Practice of Kinetic Analysis, 2nd Edition. New York, NY, USA: Wiley-Blackwell.
Wolfe RR, Shaw JH & Durkot MJ 1985 Effects of sepsis on VLDL kinetics: responses in basal state and during glucose infusion. American Journal of Physiology 248 E732–E740.
Zilversmit DB & Hughes LB 1974 Validation of a dual-isotope plasma ratio method for measurement of cholesterol absorption in rats. Journal of Lipid Research 15 465–473.
