Contents - Previous - Next

Nutrition of immune cells: The implications for whole body metabolism


* Cellular Nutrition Research Group, Department of Biochemistry, University of Oxford, South Parks Road, Oxford OX1 3QU, U.K.

1. Introduction
2. An introduction to metabolic-control logic and its application to the structure of a biochemical pathway
3. Use of maximum activities of enzymes as quantitative indices of maximum flux through metabolic pathways
4. Enzyme activities as indication of the capacity of major energy providing pathways in immune cells
5. Glutamine and the immune cells
6. Glutamine - A link between muscle and the immune system
7. Large decreases in the concentration of glutamine in plasma
8. The clinical significance of the role of glutamine in immune cells
9. The effects of glutamine provision for the patient
10. Branched-point sensitivity, substrate cycles and thermogenesis

1. Introduction

1.1. Lymphocytes
1.2. Macrophages

Lymphocytes and macrophages play an important role both qualitatively and quantitatively in the immune response, during which these cells undergo increased rates of production and recruitment, and alterations in function. It was for this reason that the nutritional requirements of these cells were investigated. Mature lymphocytes recirculate through Lymphoid tissue via blood and lymph in a relatively quiescent state until stimulated to proliferate during, for example, a bacterial or viral infection. By contrast, macrophages are terminally differentiated end-cells in which the ability to proliferate is gradually lost.

Despite the undoubted importance of these cells, it was surprising that, until the recent work from the author's laboratory, relatively little was known about their metabolism, the fuels they require to carry out their functions, the rates of utilization and the fates of these fuels, the means of control of such rates, and whether any differences in the nutrition of these cells can be related to their specific cell biology as part of the immune system. Furthermore, if these cells have unique nutritional requirements, does this have specific consequences for the animal as a whole?

1.1. Lymphocytes

Lymphocytes are small round cells with characteristic large nuclei and a thin rim of cytoplasm. Cells of similar morphology are found in the spleen, bone marrow, lymph nodes, thymus, and other areas, such as Peyer's patches and tonsils. Lymphocytes in the tissues are in dynamic equilibrium with circulating blood. They may be categorized according to their surface markers, reactions to stimuli, migratory patterns and life span. Of possible particular importance to their nutrition are their locations and relationships with other tissues in the body.

Differentiated lymphocytes originate from primitive stem cells, which in mammals reside in the bone marrow. Further differentiation leads to the development of Typhoid stem cells that are directed to specific sites of the body for further differentiation. For example, in the thymus these cells acquire certain characteristics by which they become known as T-lymphocytes, that is, thymus-derived lymphocytes; these cells take part in cell-mediated immune reactions. B-Lymphocytes mature in the bone marrow, and are the precursors of antibody-forming cells. T-cells do not synthesize detectable amounts of immunoglubulin, but function as regulators of the immune response. This is achieved via interactions between various T-cell subsets (helper, suppressor or cytotoxic T-cells) and macrophages during the operation of the cell-mediated immune response. Possible differences in nutrition and metabolism between T-cell subsets have not been studied; it is considered that any differences will be minor, but this area awaits further investigation.

In normal animals, not subjected to a serious immune challenge, most of the lymphocytes will be in a quiescent state - known as resting lymphocytes. This has been taken to mean that such lymphocytes are also metabolically inactive. However, the role of these cells in the immune system implies that they must be able, at any time, to respond rapidly and effectively to an immune challenge. This demands certain nutritional and metabolic adaptations that can only be appreciated by the knowledge of metabolic-control logic; an introduction to this topic is given below. Such adaptations and their role in the nutritional requirements of these cells may account for several of the responses to injury, trauma, sepsis, and burns and may explain, in some situations, failure of the immune system.

1.2. Macrophages

Macrophages are mononuclear cells which are formed from monocytes. They are terminally differentiated end-cells in which the ability to proliferate is gradually lost. Mature resident macrophages are found to be more widely distributed in hematopoietic, Typhoid and other tissues, where they exist as biosynthetically active cells. After an inflammatory or immunological stimulus, newly recruited macrophages with markedly different secretory and endocytic properties can accumulate in large numbers at specific sites. In spite of this functional heterogeneity, macrophages overall are characterized as motile, highly phagocytic cells that display marked plasma-membrane activity. Their versatile secretory activities, which include enzymes, hormone-like mediators of inflammation, and toxic oxygen products can be altered by stimuli, such as phagocytic particles, microbial products and lymphokines acting on various macrophage plasma-membrane receptors. They have, therefore, a high turnover of mRNA to provide for the synthesis of different secretory proteins. The cell-biological function of these cells within the immune system would, therefore, appear to be quite distinct. It would not be expected a priori that their nutritional requirements and fuel metabolism, at least those that relate to cell function, would be similar; the evidence provided below, however, is that they are.

The first question is how is it possible to approach the question of which fuels particular cells might use in vivo, and at what rates. The answer requires some knowledge of metabolic-control logic which is given briefly below.

2. An introduction to metabolic-control logic and its application to the structure of a biochemical pathway

2.1. Near-equilibrium and non-equilibrium reactions
2.2. The flux-generating reaction

2.1. Near-equilibrium and non-equilibrium reactions

Reactions in a metabolic pathway can be divided into two classes: those that are very close to equilibrium (near-equilibrium) and those that are far removed from equilibrium (non-equilibrium). For a non-equilibrium reaction, the rate of the reverse component (Vr) of the reaction is very much less than the rate of the forward component (Vf), for example:

Example 1

One important point emerges from this: unless it has a kinetic effect, a change in concentration of the product will have only a negligible effect on the rate of the reverse reaction and hence on the net rate. With a near-equilibrium reaction, the rates of the forward and the reverse components of the reaction are much greater than the overall net rate and are similar to one another, for example:

Example 2

In this reaction, change in the concentration of the product now has a marked 'mass-action' effect on the rate of the reverse reaction and hence the net rate.

The equilibrium nature of a reaction can be decided on the basis of the free energy change of the reaction (D G), which can be calculated from a knowledge of the ratio of the concentration of product to the concentration of substrate in the living tissue, known as the mass action ratio, or G and the equilibrium constant Keq, of the reaction as follows:

D G = - RTIn Keq + RTIn G

When D G is <1.0 kcal/mol, the reaction is considered to be non-equilibrium. A problem arises in that there is a 'grey' area, close to 1 kcal/mol in which it is difficult to decide upon the equilibrium nature (NEWSHOLME and CRABTREE, 1976).

2.2. The flux-generating reaction

If an enzyme catalyses a non-equilibrium reaction in a biochemical process and approaches saturation with its pathway-substrate (that substrate which represents the flow of matter through the pathway), so that the catalytic rate is largely independent of that substrate concentration, the reaction can be regarded as the flux-generating step for the pathway. In other words, in the steady-state, this reaction initiates a flux to which all the other reactions in the pathway must adjust. (Such a reaction must approach saturation with its pathway-substrate since, if it did not, as the reaction proceeded the substrate concentration would decrease and this would decrease the rate of the reaction and hence the flux through the pathway; a steady-state would then be impossible.) One important development from the concept of a flux-generating step is that it provides a physiologically useful definition of a metabolic pathway. A pathway is defined as a series of reactions that is initiated by a flux-generating step and ends with the loss of end-product to a metabolic sink (storage form) or to the environment; or it ends in a reaction that precedes another flux-generating step. The advantage of this definition is that it allows us to discuss more objectively the functions of processes and to define limitations in understanding these functions. This will be illustrated by reference in particular to the process of glutamine utilization by immune cells. The questions, what is the importance of this information on metabolic control logic and how can we obtain information about fuel utilization in cells, are now answered.

3. Use of maximum activities of enzymes as quantitative indices of maximum flux through metabolic pathways

The advantage of a near-equilibrium reaction, in a metabolic pathway in vivo, is that the reaction may be very sensitive to small changes in concentrations of cosubstrate or coproduct. Consequently, large changes in flux can be transmitted through such a reaction without any requirement for complex regulatory properties. In general, this means that the activity of the enzyme can be measured relatively easily in crude extracts of the tissue; this ease of assay has, unfortunately, been used by some investigators as the only criterion for the selection of an enzyme to study the maximum flux through a metabolic pathway. This cannot be done. Metabolic logic tells us that these enzymes cannot be used as quantitative indices of flux but, despite this, they are still being used that way.

Enzymes that catalyse non-equilibrium reactions in a metabolic pathway provide directionality in that pathway and are usually subject to allosteric control (see above). Indeed, the control mechanisms may be complex. This means that knowledge of such control mechanisms must be available before a satisfactory assay method for measurement of the maximum activity can be developed; hence, knowledge of metabolic control is necessary to enable the enzyme activity to be adequately assayed in crude extracts of the tissue (see below).

There are at least two conditions that must be satisfied before an enzyme activity can be used to provide quantitative information. First, it is necessary to establish which enzymes in the pathway catalyse non-equilibrium reactions (see above). Second, it is necessary to demonstrate experimentally that the maximum activities of such enzymes in vitro can be used to indicate quantitatively the maximum flux through a reaction. This is done by comparison of the in vitro enzyme activity with the measured or calculated maximum flux through the pathway (NEWSHOLME and LEECH, 1983).

4. Enzyme activities as indication of the capacity of major energy providing pathways in immune cells

Systematic studies in the 1970s on the maximum activities of key enzymes of carbohydrate and fat metabolism in muscle provided information on the types of fuel utilized by different muscles and their maximum contribution to ATP formation to support contractile activity. This enabled a systematic and comprehensive analysis to be made of the fuels used by different muscles from different animals across the animal kingdom. This knowledge together with increasing quantitative information on amounts of fuels in muscle and interaction between tissues has led to new ideas on causes of fatigue and limitations in performance in Olympic events, and these may apply to fatigue caused by undernutrition, injury or viral infections.

More recently, a similar approach has been applied to lymphocytes, macrophages, fibroblasts and endothelial cells. This has provided, for the first time, evidence that these cells can use glutamine and/or long-chain fatty acids for energy formation and, indeed, that these fuels could be quantitatively more important than glucose.

5. Glutamine and the immune cells

Three lines of evidence suggest that glutamine is used at a very high rate by lymphocytes and macrophages:

1. The maximal catalytic activities of a number of key enzymes in the metabolism of glutamine - including glutaminase - have been measured in lymphocytes and in macrophages, and they are high (ARDAWI and NEWSHOLME, 1983; 1985).

2. Rates of utilization of glutamine by isolated lymphocytes and by macrophages incubated for 60-90 minutes have been measured, and they are also high (NEWSHOLME et al., 1987).

3. The rates of utilization of glutamine by these cells in culture are similar if not higher than those of the incubated cells.

Interestingly, very little of the carbon of glucose and not very much of the carbon of glutamine is oxidized via acetyl-CoA and the classic Krebs cycle, despite the fact that enzymes of the full Krebs cycle are available in these cells. Glucose is converted almost totally into lactate, glutamine into glutamate, aspartate and lactate, and fatty acids into ketone bodies and possibly other end-products (ARDAWI and NEWSHOLME, 1985; NEWSHOLME et al., 1988).

Before, it had generally been assumed that both lymphocytes and macrophages obtained most of their energy from glucose, and, furthermore, that lymphocytes which had not been subjected to an immune response (resting or quiescent lymphocytes were metabolically inactive. A simple comparison shows the naively of such a view: the rate of utilization of glucose and glutamine by resting lymphocytes is about 25% of the rate of glucose utilized by the maximally physically working heart muscle (Table 1.).

The question arises, therefore, as to the significance of these high rates. The role of partial oxidation of glutamine (glutaminolysis) in rapidly dividing cells has been considered previously to be the provision of energy and/or the provision of both nitrogen and carbon for precursors for synthesis of macromolecules (e.g., purine and pyrimide nucleotides for DNA and RNA). However, there are problems with both explanations. Quantitative information, provided by the work from the author's laboratory with lymphocytes shows that the rate of glutaminolysis is markedly in excess of that of the precursor requirements; and the actual rate of glutamine utilization is at least 30-fold greater than the maximum capacity for pyrimidine, and therefore presumably purine, nucleotide synthesis (Table 2.; SZONDY and NEWSHOLME, 1989). And, since all the enzymes of the Krebs cycle appear to be present in these cells (ARDAWI and NEWSHOLME, 1985; NEWSHOLME et al., 1988), if energy generation per se was the major reason for high rates of glutamine utilization, it would be expected that more of the carbon skeleton of glutamine would be converted to acetyl-CoA for complete oxidation via the Krebs cycle (i.e., that glutamine oxidation rather than glutaminolysis would occur).

So why is glutamine used at such a high rate? On the basis of metabolic-control logic (CRABTREE and NEWSHOLME, 1985), it has been suggested that the high rates of glutaminolysis (and glycolysis) provide optimal conditions for the precise and sensitive control of the rate of use of the intermediates of these pathways for biosynthesis precisely at the time required by the synthetic processes during the cell cycle (e.g., glutamine for purine and pyrimidine nucleotide synthesis [NEWSHOLME et al., 1985a; b]). Consequently, any decrease in the flux through this glutamine pathway - even if small - could impair the functioning of the immune system (NEWSHOLME et al., 1988).

Table 1. Rates of utilisation of glucose or glutamine by lymphocytes and macrophages and other tissues in mouse, rat and man

Data from NEWSHOLME and LEECH (1983) and ARDAWI and NEWSHOLME (1985). It should be noted that the rate of utilisation of glutamine by lymphocytes in man may be equivalent to the rate of glucose utilisation by the brain - and since the amount of tissue is approximately the same (c. 1400 g) the demand for glutamine by cells of the immune system may be similar to that of the brain for glucose.



Rates of utilisation (nmol/h) per mg protein








Mesenteric lymphocytes



Maximally-working heart




Peripheral lymphocytes






Table 2. Maximal activity of carbamoylphosphate synthetase II (CPS-II) (capacity of pyrimidine nucleotide synthesis), rates of synthesis of uridine nucleotides and rate of glutamine utilisation by lymphocytes

Data are taken from SZONDY and NEWSHOLME (1989). The activity of CPS-II represents the maximum capacity for pyrimidine nucleotide synthesis in these cells, which is considerably less than the actual measured rate of uridine nucleotide synthesis under the conditions of the experiment.

Rates (nmol/h per mg protein) Capacity for pyrimidine nucleotide synthesis (CPS-II activity)

Actual rate of uridine nucleotide synthesis [14C]-incorporation from HCO3 into uridine nucleotides

Actual rate of glutamine utilisation




A decrease in the level of glutamine in culture has been shown to markedly decrease the proliferation of rat mesenteric lymphocytes (SZONDY and NEWSHOLME, 1989) and human lymphocytes (PARRY-BILLINGS et al., 1990); and a decrease in the concentration of glutamine has also been shown to decrease the rate of phagocytosis in macrophages (PARRY-BILLINGS et al.,, 1990). It is important to appreciate that the high rate of glutamine utilization (as well as glucose and fatty acids) and its partial oxidation, plus the effects of decreases in the concentration of glutamine, are predicted by the theoretical work on the problems of providing precision in metabolic control by CRAB TREE and NEWSHOLME (1985).

In non-mathematical terms, high sensitivity is achieved because the rates of the biosynthetic pathway (e.g., purine or pyrimidine nucleotide synthesis) can be increased markedly without decreasing significantly the concentration of the metabolic precursors (e.g., glutamine, aspartate) which are constituents of the main pathway. If a decrease in concentration of glutamine, for example, occurred, it would 'oppose' the stimulation of the biosynthetic pathway, and the precision in the provision of the correct increase in the concentrations of nucleotides for the cell cycle would be lost. Branched-point sensitivity allows the maintenance of a precise increase in the rate of biosynthesis of key compounds to occur. Thus, in lymphocytes, this high sensitivity will be particularly important when the macromolecular synthesis is required during proliferation in response to an immune challenge.

In the terminally differentiated macrophage it is likely to be for synthesis of mRNA for producing secretory and receptor proteins and enzymes, when demanded by an immune challenge or in response to injury or burns (NEWSHOLME et al., 1988). Furthermore, this explanation provides an answer to the question of why these cells do not appear to allow pyruvate, generated from either glucose or glutamine, or acetyl-CoA generated from b-oxidation of fatty acids, to be oxidised. If oxidised, the pyruvate or acetyl-CoA would provide a large amount of energy (via the Krebs cycle and electron-transport process) and, since a high ATP concentration inhibits key reactions in both glycolysis (e.g., 6-phosphofructokinase) and glutaminolysis (e.g., oxoglutarate dehydrogenase), high rates of these latter processes would not be possible; the advantage for control of rates of biosynthesis would be lost. It is, however, possible that, if at some stage of the cell cycle the demand for ATP becomes very large, the rate of the complete Krebs cycle will increase dramatically. The principle of branched-point sensitivity may, therefore, be of considerable importance in understanding some unusual characteristics of those cells with specific and key responsibilities in the immune system.

The important point to emerge from this discussion is that glutamine must be used at a high rate for the cells of the immune system even when they are quiescent. The immune response to invasion by a microorganism must be rapid; hence, the potential rate of glutamine utilization must always be high to provide optimal conditions for response to an immune challenge at any time. A decrease in the concentration of glutamine available to these cells could seriously impair the functioning of the immune system with all that this would imply. This then raises the question as to the source of this glutamine?

6. Glutamine - A link between muscle and the immune system

6.1. Glutamine synthesis in skeletal muscle
6.2. The transport of glutamine across the muscle membrane: Glutamine uptake and release

Both liver and muscle can produce and release glutamine into the bloodstream, but muscle may be quantitatively the most important tissue. Since the cells of the intestine have a large capacity to utilize glutamine, most of the glutamine that enters the body via the diet is utilized by the intestine. Hence glutamine production and release by muscle become of considerable physiological and immunological importance.

The systematic study of glutamine flux in skeletal muscle, using the principles of metabolic-control logic, has highlighted two areas of potential misunderstanding in the literature. These relate to the processes of glutamine synthesis in muscle and the processes of glutamine uptake and release across the muscle membrane. It should be emphasized that glutamine synthesis, uptake and release by muscle are three distinct processes. Evidence has been obtained that the process of glutamine release from skeletal muscle may be the limiting step - indeed may be the flux-generating step - for maintenance of the plasma glutamine level and hence its uptake by other cells (e.g., immune cells).

6.1. Glutamine synthesis in skeletal muscle

The terms glutamine 'synthesis' or 'production' and glutamine 'release' from muscle have not always been used accurately and systematically in the literature. For example, it has been assumed that the rate of glutamine release is equivalent to that of glutamine synthesis by skeletal muscle, and vice versa. This has led to some confusion, since glutamine synthesis and release from muscle appear to be independent processes. Studies of the process of glutamine synthesis usually involve the measurement of the maximal activity of the enzyme which catalyses the synthesis of glutamine (glutamine synthetase) and the determination of the kinetic properties of this enzyme in vitro.

The question arises as to whether the synthesis of glutamine is important in the control of glutamine release from skeletal muscle and hence in the metabolism of this amino acid in a number of tissues, including cells of the immune system. Given the high concentration of glutamine in skeletal muscle (approximately 20mM in man, 8mM in the rat), it is unlikely that the synthesis of glutamine in muscle can limit or control the rate of glutamine release from muscle, at least under physiological conditions. Thus, although the activity of glutamine synthetase may be increased in conditions associated with an increase in the rate of muscle glutamine release (e.g., burns, cancer cachexia, dexamethasone treatment), this enhancement of glutamine synthesis may function to maintain a steady-state concentration of glutamine in muscle, but not to control the flux through the pathway.

The importance of the synthesis of glutamine in muscle, for the maintenance of a steady-state concentration of glutamine, is high-lighted by the following calculation in the rat. Assuming muscle glutamine concentration is 4000 nmoles/g and the rate of muscle glutamine release is 40 nmoles/min/g, it may be calculated that the total skeletal muscle glutamine pool of a rat could be depleted within 100 minutes, if no synthesis of glutamine occurred. However, this does not mean that synthesis is responsible for the control of release. This appears to be the role of the efflux process across the plasma membrane (PARRY-BILLINGS et al., 1989). The possibility, nevertheless, arises that the rate of glutamine synthesis may limit glutamine release from muscle in conditions associated with a marked depletion of the muscle glutamine pool, as may occur in some pathological conditions.

Thus, although glutamine synthesis may be important in certain pathological conditions, the release of glutamine from muscle, and not glutamine synthesis, is likely to be the flux-generating step. It follows, that in order to understand the factors which affect the plasma concentration of glutamine, attention should be focused particularly on the control of muscle glutamine release, rather than on its synthesis.

6.2. The transport of glutamine across the muscle membrane: Glutamine uptake and release

Glutamine may be both released and taken up by skeletal muscle (RENNIE et al., 1986). Glutamine uptake is a sodium-dependent transport process. Unfortunately, it appears to have been tacitly assumed that the inward transport of glutamine is the same process as that of glutamine release from muscle. It does not appear to have been considered that uptake and release could be two separate processes. This has perhaps led to some confusion over the release process. RENNIE et al. (1986) suggested that glutamine transport across the membrane was 'symmetrical'. This would imply that the process was near-equilibrium. This does not, however, appear to be the case.

Calculation of the free energy change for glutamine uptake suggests that it may be a non-equilibrium process and therefore may not be. 'symmetrical' (D G is approximately 2.94 kcal (12.3 kJ)/mole). This is also the case if the efflux process is flux-generating; such processes cannot be near-equilibrium. Thus, if this calculation is correct, glutamine uptake and release by muscle would be different processes, and the release of glutamine would involve a different transporter. This transporter could possess different kinetic properties than that involved in glutamine uptake and could be controlled by totally different regulators. It has been calculated that the Km for glutamine of the release process must be substantially lower than 4mM and is thus different from the Km determined for the process of glutamine uptake (i.e., 7mM; RENNIE et al., 1986). However, even if the processes of glutamine uptake and release are two separate and in dependent processes, it does not mean that the uptake process has no role in the control of glutamine release.

If muscle glutamine release and uptake are two separate processes, which are both non-equilibrium and which 'catalyse' the same pathway reaction in opposite directions, then the possibility arises that a substrate cycle exists between extracellular and intracellular glutamine pools. A substrate cycle is produced when a non-equilibrium reaction in the forward direction is opposed by another non-equilibrium reaction in the reverse direction of the pathway (NEWSHOLME and CRABTREE, 1976). Thus, in the present discussion, muscle glutamine release and muscle glutamine uptake would constitute the forward and reverse reactions of the substrate cycle, respectively.

This proposed cycle may also be termed a 'translocation' cycle, since it does not involve the chemical interconversion of two different metabolites (see NEWSHOLME and CRABTREE, 1976). An analogy may be drawn with the proposed translocation cycle involving sodium ion entry into cells and the sodium 'pump' process. A further analogy may be drawn with the proposed substrate cycle for the transport of calcium ions across the mitochondrial membrane. Calcium release and uptake by mitochondria are mediated by sodium-dependent and independent transport processes, respectively (CROMPTON et al., 1976).

An important implication of the existence of a substrate cycle for the release of glutamine from skeletal muscle is that substrate cycling could enhance the sensitivity of metabolic regulation. Indeed, given the proposed importance of glutamine release from skeletal muscle, it is not surprising that the sensitivity of control of this process should be enhanced by such a mechanism.

7. Large decreases in the concentration of glutamine in plasma

The plasma concentration of glutamine is decreased in a number of conditions. The important consequence of a decrease in the concentration of glutamine in plasma of both rat and man may be an impairment of immune function (see above). Furthermore, since immune deficiency may be associated with an increased incidence of clinical complications and mortality after major injury (GREEN and FAIST, 1988), it is suggested that decreases in the plasma glutamine concentration may be closely involved in some of the problems of some pathological conditions.

Recent studies have shown that the plasma level of glutamine is dramatically decreased in patients with severe burns and is decreased less severely in subjects after a marathon (PARRY-BILLINGS et al., 1990). The immune system is known to be suppressed under these conditions (see above). These are the first studies to demonstrate such marked decreases in the plasma glutamine levels in these conditions and provide, at least in part, a possible explanation for the immunosuppression that characterizes these conditions. Indeed, this is the first work to show such a massive decrease in the plasma glutamine level in any condition. The plasma glutamine level may be dramatically decreased in other conditions (major surgery, severe sepsis, major trauma) and the response of lymphocytes and perhaps macrophages to an immune challenge may then be decreased. This is a hypothesis that should be tested in the future.

Contents - Previous - Next