Contents - Previous - Next

Chapter 7: The effect of isotope sequestration and exchange

7.1 Isotope incorporation into tissue and its effect on the DLW technique
7.2 The stoichiometry of sequestration
7.3 The effect of sequestration on the performance of DLW
7.4 Export of exchangeable hydrogens
7.5 The effect of sequestration/exchange on related techniques
7.6 Summary and conclusions
7.7 References

Contributor: Paul Haggarty

7.1 Isotope incorporation into tissue and its effect on the DLW technique

This chapter sets out to describe the process of isotope sequestration and the nature of its effect on the DLW method; to present appropriate constants and numerical methods whereby the effect of sequestration may be calculated; to outline the evidence that sequestration occurs in human subjects; and to assess the likely limits of error in normal human subjects and in special groups where the problem of sequestration is exacerbated.

7.1.1 Labile and non-labile hydrogen

Label movement into and out of products other than CO₂ and water may be reduced to two distinct processes:

a) exchange of water hydrogen and oxygen with labile hydrogen and oxygen;
b) incorporation of water hydrogen and oxygen into, and release from, non-labile positions.

Labile hydrogens are found on carboxyl, hydroxyl, amino and sulphydryl groups ¹, and non-labile hydrogens are those bonded to carbon ². Exchange is a rapid process but sequestration (ie incorporation of water hydrogen into, and release from, non-labile positions) occurs over longer periods. It is important to distinguish between these processes because they are dealt with in different ways in the DLW technique.

7.1.2 Exchange

The most obvious manifestation of the exchange process is found in the difference between body water and deuterium and oxygen-18 distribution spaces. This difference occurs because the ²H₂O and H₂¹⁸O distribute throughout the body water, but deuterium can also exchange with the labile hydrogens in body solids resulting in a dilution space greater than the body water pool.

Exchange of ¹⁸O with bicarbonate also occurs, indeed the DLW technique is based on this fact, therefore the ¹⁸O distribution space will also be greater than the body water space. Taking values derived by Irving et al ³ for the bicarbonate pool-size in man (11300 umol/kg body weight), together with the assumption that ¹⁸O is in equilibrium with this pool and that 1 mole of bicarbonate can exchange oxygens with 3 moles of water it follows that the bicarbonate pool is equivalent to 0.6105 g water/kg body weight. Thus in a 50 kg subject with a body water pool size of 28 kg the ¹⁸O distribution space will over-estimate body water by 30.525 g or 0.1%. In reality it has been observed that H₂¹⁸O dilution over-estimates body water by approximately 1% suggesting that there are other exchange reactions involving ¹⁸O.

Whatever factors are involved in the exchange processes the only measurable effect on the DEW technique is to increase the isotope distribution spaces. The use of independent pool sizes ^{4, 5} which are not necessarily equivalent to body water, is sufficient to deal with this phenomenon, at least with respect to exchange within the body. Exchange with material which is subsequently exported from the body (eg urine, faeces or milk) will for all practical purposes appear as sequestration and will therefore not be adequately dealt with by using independent pool sizes.

7.1.3 Sequestration

Deuterium incorporation into body solids during reductive biosynthesis creates the greatest concern with respect to sequestration. The term 'reductive' is used to describe any process whereby oxygen is removed from a compound and/or hydrogen is added. In many cases the added hydrogen has its origin in water. This is not surprising since water is the universal solvent for chemical reactions in the body and the hydrogen of water is rapidly exchangeable with many of the intermediates involved in the biosynthetic process. That sequestration occurs is well documented, and H isotope incorporation from water into protein and fat has even been used to measure turnover rates in vivo ^{6, 7}. Unlike exchange, where the simple use of independent pool sizes is sufficient to deal with its effect on the calculation of r_CO2, the process of sequestration can only be dealt with by estimating the rate of hydrogen isotope incorporation and by correcting the derived parameters of r_H2O and r_CO2 by the appropriate amount.

Water hydrogen is known to be incorporated into stable bonds but the important questions are: Does it occur to such an extent that it will have a measurable effect on the technique? If so, what is the likely magnitude of error? In order to answer these questions we must first obtain information on the stoichiometry of water hydrogen incorporation into the main constituents of the body and second, we must know the rate of synthesis of these constituents.

7.2 The stoichiometry of sequestration

7.2.1 Fat

Fat is used here to describe free fatty acids, mono-, di-and tri-glycerides. The fatty acids which form the largest part of the glycerides are synthesised from 2-carbon units by the enzymes acetyl CoA carboxylase and fatty acid synthetase. The reductant in this biosynthetic process is NADPH and the hydrogen in water is freely exchangeable with the NADPH hydrogen, thereby providing the route of entry of water H into the stable carbon-hydrogen bonds of the fatty acid molecule. The stoichiometry of incorporation can be deduced from the empirical observations of early workers in the field who maintained mice on a constant intake of deuterated water over a period of weeks and then measured the deuterium content of the body fat. Only the saturated fats were studied and it was observed that 43 - 46% of the hydrogen atoms of newly synthesised fatty acids were derived from body water ^{2, 8-10}. However, it has subsequently been pointed out by Jungas ² that the close agreement between these studies was fortuitous since the time interval used (21-98 days) "did not allow for the de novo synthesis of the entire carcass pool of saturated fatty acid" but that approximately "53% of the hydrogens of newly synthesised fatty acids would be derived from water as assayed with deuterium". Jungas was careful to specify deuterium because mass effects will cause the factor to be different for tritium. In studies where both isotopes were used Jungas observed that 17% less ³H than ²H was incorporated into newly synthesised fat. For the purposes of the calculation of sequestration the value of 53% will be used for the stoichiometry of water H isotope incorporation when assayed with ²H and 44% when assayed with ³H.

In order to calculate the extent of incorporation we must also know the composition of synthesised fat. From data on the composition of human adipose tissue the average fatty acid formula can be calculated as C_17.13H_31.65O (molecular weight 254 a.m.u.) for normal adult humans 11. This gross composition changes very little between males and females or in pathological conditions such as coronary heart disease, diabetes and obesity. The empirical formula for the triglyceride of this fatty acid is C_54.41H_99.97O₆ (molecular weight 850 a.m.u.). Assuming that 53% of the hydrogens of newly synthesised fatty acid will be derived from water labelled with ²H, synthesis of 1 mole of fatty acid will result in the incorporation of 16.78 moles of ²H and deposition in the form of triglyceride will result in the sequestration of 50.34 moles of ²H per mole of triglyceride, or 59.19 mmoles per gram. Were incorporation to be assayed with tritiated water the values would be 13.93, 41.78, and 49.13 respectively. In terms of equivalent water flux this translates into 0.5333 g of water/g fat synthesised (using ²H₂O) and 0.4427 g of water/g fat synthesised (using ³H₂O).

7.2.2 Protein

Similar experiments to those described for fat have also been carried out, largely by the same workers, to study the incorporation of water H into the stable carbon-hydrogen bonds of the amino acids which go to make up the body proteins. Unlike fat, however, there are a large number of exchangeable hydrogens in protein ¹. Early workers therefore took great care to remove these by washing before analysing the amino acids for deuterium in stable carbon-hydrogen bonds. Foster, Rittenberg and Schoenheimer ¹² measured the incorporation of ²H from water into non-labile positions in nine amino acids of the hydrolysed protein of rats maintained on deuterated water for 10-19 days and found that, on average, 0.9 atoms of ²H derived from water were incorporated into stable carbon-hydrogen bonds per amino acid. The hydrogen of water is thought to be incorporated into the a-carbon of the amino acid during transamination ⁷. In studies on transamination in vitro, Hilton et al ¹³ found that 1.7 deuterium atoms were incorporated per molecule of aspartate. More recently Commerford et al ¹⁴ performed similar in vivo experiments where mice were labelled with tritiated water from conception and looking at 16 amino acids they found that, on average, 0.41 atoms of heavy hydrogen were incorporated per amino acid. Using the figure of 17% less incorporation for ³H than ²H (derived for fat synthesis), we can calculate that approximately 0.5 atoms of ²H from water would be incorporated per amino acid. The discrepancy between the values of 0.9 and 0.5 may partly be explained by the smaller number of amino acids used by Foster et al, since if Commerford et al's analysis is restricted to the same amino acids the average is closer to 0.6. For the purposes of the calculation of sequestration the more conservative figure of 0. 5, based on a larger sample of amino acids, will be used for water labelled with ²H and 0.4 for water labelled with ³H.

In order to calculate the extent of incorporation we must also know the composition of synthesised protein. The main proteins in the human body are collagen, actin, tropomysin and myosin which respectively have 0. 8965, 0.7919, 0.8133 and 0.8227 mmols of amino acids per gram of protein ¹. Since these values are very similar, and in the absence of any knowledge of the respective rates of synthesis of each in the body, an average of 0.8227 mmols of amino acids per gram of synthesised protein will be used. Using this and the values for water H incorporation we can calculate that 0.4114 mmols will be incorporated per gram of protein when assayed with ²H and 0. 3291 mmols when the isotope is ³H. In terms of water flux this represents 0.003706 g water/g protein synthesised (using ²H₂O) and 0.002965 g water/g protein synthesised (using ³H₂O).

7.2.3 carbohydrate

The main storage form of carbohydrate in the body is glycogen and the stoichiometry of water hydrogen incorporation into liver glygogen has again been studied in small animals using deuterated water. Stetten and Klein ¹⁵ observed that about 38% of the hydrogen in liver glycogen arose from the body water when assayed with deuterium and that this deuterium was uniformly distributed among the non-exchangeable positions. This value varies slightly depending on the precursors available but 38% is a good approximation. Carbohydrate, like protein, has a large number of exchangeable hydrogens therefore care has to be taken to remove these before estimating the carbon bound deuterium. Stetten and Boxer ¹⁶ found that 34% of the hydrogens of glycogen are freely exchangeable with water when the label is ²H. From the chemical formula of glycogen it can be calculated that 22.33 mmoles of ²H derived from water will be sequestered per gram of glycogen synthesised This translates into 0.2012 g water/g glycogen synthesised (using ²H₂O) and (using the ³H/²H mass discrimination effect observed during glycogen synthesis of 8% ¹⁷) 0.1851 g water/g glycogen synthesised (using ³H₂O).

For ease of reference the values derived above are summarised in Table 7.1.

Table 7.1. Water equivalents incorporated into fat, protein and carbohydrate (assayed with ²H₂O)

^a As triglyceride: chemical formula C_54.11H_99.97O₆

^b Consisting of an equal mixture of collagen, actin, tropomyosin and myosin.

^c As glycogen.

Contents - Previous - Next