Gut 1994; supplement 1: S13-S17 81S13 Quantitive aspects of glucose and glutamine

metabolism by intestinal cells Gut: first published as 10.1136/gut.35.1_Suppl.S13 on 1 January 1994. Downloaded from

E A Newsholme, A-L Carrie

Abstract Fuel utilisation in the intestine Gut fuel utilisation has several unique Fuels are required to provide energy for the features. Arterial and luminal fuels cell. According to conditions, various fuels provide nutrition for the enterocyte, the (glucose, triglycerides, fatty acids, ketone former being of more importance. This bodies, and amino acids) are available both in factor, and the heterogeneity of cell types the bloodstream and in the intestinal lumen. within the gut makes it difficult to define As fuels can be supplied to the intestine from its fuel utilisation. Metabolic control logic the vascular bed or from the lumen, the suggests that modulation of the maximal physiological significance of the substrate will activity of any pathway resides in those be different. In general, substrates absorbed enzymes that operate in vivo at rates far from the lumen are eventually released into the below their maximal capacity and that blood (as parent compound or metabolites) catalyse non-equilibrium reactions. On and distributed to other tissues. In contrast, this basis, although enterocyte hexokinase the utilisation of substrates from blood is prob- activity is much higher than in other ably followed by an extensive metabolism in 'glycolytic' cells (for example, brain), the intestine. Even if substrates from the lumen potentially high rates ofglucose utilisation can be used to provide energy, the intermit- are modulated by substrate cycling of tency of alimentation will still require an glucose 6-phosphate back to glucose adequate vascular energy supply, especially through glucose 6-phosphatase. Gluta- between meals. Therefore a distinction must mine metabolism proceeds by glutaminase be made between the terms 'substrate uptake' to produce glutamate, which may then and 'substrate utilisation'. The first refers to be transaminated (aspartate-aminotrans- the disappearance of the substrate from the ferase and alanine-amino transferase) to medium (for example, vascular or luminal produce a-ketoglutarate, alanine, and perfusates), whereas the second refers to the aspartate. The end products of glutamine difference between the amount of substrate metabolism by incubated gut preparations absorbed - that is, from the lumen - by the in vitro (mainly alanine), suggests that intestine and that which is released on the http://gut.bmj.com/ enterocytes, not immune cells, are respon- other side - that is, into the blood. This differ- sible for most gut glutamine metabolism. ence is not always made clear - possibly High flux rates of glucose and glutamine because it is not easy to distinguish between metabolism in the enterocyte may result the two.' Two important questions, therefore, from the need for de novo synthesis of are which fuels are used by the intestine, and purines and pyrimidines and ribose sugars whether the pattern of fuel utilisation or the for nucleic acid synthesis. Sepsis reduces rate of utilisation, or both depend on the site of on September 23, 2021 by guest. Protected copyright. rates of glucose and glutamine metabo- uptake of the fuel? Unfortunately, available lism, perhaps to preserve the increased data do not permit complete answers to these consumption of these fuels by activated questions, but it is nevertheless necessary to lymphocytes and macrophages in the gut understand how enzyme activities in the test wall. tube can provide useful information about fuel (Gut 1994; supplement 1: S13-S 17) utilisation in vivo.

In this paper, aspects of fuel utilisation by the Use of maximum activities of enzymes as small intestine are compared with those in quantitative indices of maximum flux other tissues, in an attempt to raise important through metabolic pathways questions concerning the nutrition of the gut. The theory underlying the use of maximal Firstly, the interrelation between glucose and enzyme activities to show the maximum Cellular Nutrition glutamine utilisation by the intestine is dis- capacity for fluxes in biochemical systems has Research Group, cussed. The role of enterocytes as contributors been described elsewhere.' Enzymes can be Departnent of to the whole small intestinal glutamine metab- classified as to whether they catalyse reactions Biochemistry, olism has been established from the metabolic far removed from University of Oxford, equilibrium (non-equilib- Oxford characteristics of enterocytes compared with rium) or near equilibrium. The advantage of E A Newsholme, those of other cells that constitute the intestinal monitoring a near equilibrium reaction in a A-L Carrie mucosa (for example, lymphocytes). In addi- metabolic pathway in vivo is that the reaction Correspondence to: tion, a hypothesis for the regulation of glucose may be very sensitive to small changes in Dr E A Newsholme, Cellular metabolism in the intestine and its concentrations or Nutrition Research Group, possible of cosubstrate coproduct. Department of Biochemistry, significance for the whole animal is put Consequently, large changes in flux can be University of Oxford, South forward. the effects of on Parks Road, Oxford Secondly, sepsis transmitted through such a reaction without OX1 3QU. intestinal metabolism are briefly discussed. any requirement for complex regulatory S14 Newsholme, Carri

properties. In general, the activity of the TABLE II Maximalflux through glycolysisfrom glucose enzyme can be measured comparatively easily and maximal activites ofhexokinase in various tissues in crude tissue extracts; the convenience ofthis Preparation Glycolyticflux Hexokinaseflux

assay has, unfortunately, been used by some Gut: first published as 10.1136/gut.35.1_Suppl.S13 on 1 January 1994. Downloaded from Enterocytes 602 9-8 investigators as the sole criterion for selecting Lymphocytes 37-2 28-2 an enzyme to study the maximum flux through Murine macrophage 1050 10 4 a metabolic pathway. This cannot be done. Colonocytes 414 8-2 Metabolic control logic tells us that these enzymes cannot be used as quantitative indices for the first time, evidence that such cells can offlux' but despite this they are still being used use glutamine or long chain fatty acids, or both, even in the 1990s! for energy formation and, indeed, that these Enzymes that catalyse non-equilibrium fuels could be quantitatively more important reactions in a metabolic pathway provide than glucose.3-5 (It was previously believed that directionality in that pathway and are usually glucose was the main, if not the only, fuel to be subject to allosteric control. Indeed, the used by lymphocytes.) In addition, such studies control mechanisms may be complex. Thus, have provided quantitative information on the knowledge of such control mechanisms must rates of energy production from different fuels be available before a satisfactory assay for for macrophages in culture.6 This 'maximal measurement of the maximum activity can be enzyme activity approach' has been used for the developed; hence, knowledge of metabolic small intestine. control is necessary to permit maximum enzyme activity to be assayed adequately in tissue extracts. Glucose metabolism in the small intestine There are at least two conditions that must Hexokinase activity in the mucosa of the small be satisfied before enzyme activity can be used intestine is lower than that of other glycolytic to provide quantitative information about enzymes (Table I). Its activity is, however, maximum rates of fuel utilisation in vivo. more than 10-fold greater than the reported Firstly, it is necessary to establish the reactions maximal flux of glucose through glycolysis in the pathway that catalyse non-equilibrium (Table II). This suggests either that the study reactions (see above). Secondly, it is necessary conditions were not conducive to glycolysis or to show experimentally that the maximum that the flux through this enzyme is normally activities of such enzymes in vitro can be used under considerable inhibitory control within to show quantitatively the maximum flux the enterocyte (see below). through the pathway. If it is accepted that hexokinase activity Systematic studies in the 1970s examined the points to the maximum flux through glycolysis maximum activities of key enzymes of carbo- from glucose in the enterocytes, (as is the case hydrate and fat metabolism in muscle. in muscle and brain), it is clear that the

Information was gleaned on the types of fuel intestine's maximal capacity to use glucose is http://gut.bmj.com/ used by different muscles and their maximum greater than that of the brain, and yet glucose contribution to energy formation to support is known to be an obligatory and important contractile activity. This information permitted fuel for the brain (Table III). This is of con- a systematic and comprehensive analysis to be siderable importance, because the hexokinase made ofthe fuels used by different muscles from activity in the intestinal cells is sufficient to use different animals across the animal kingdom.1 2 almost all glucose consumed in a normal rat

More recently, a similar approach has been diet (assuming 60% of the energy of the diet is on September 23, 2021 by guest. Protected copyright. applied to lymphocytes, macrophages, and carbohydrate). endothelial cells. These studies have provided, A mechanism must, therefore, exist to decrease the rate of glucose utilisation within the enterocyte. One possible mechanism is TABLE I Effects ofendotoxin treatment on activities ofenzymes ofglucose metabolism the activity of glucose-6-phosphatase; its maxi- Enzyme activities mum activity is similar to that of hexokinase in the fed state (Table I). This mechanism would Conditions ,umol/min- fresh wt nmollmin- l/mg- 1 protein Enzyme I/g- require that both enzyme activities occur in the Hexokinase Fed 11-0 65-3 Starved 8-4* 48-9* same cell and that both enzymes are simul- Septic 6-2* 47-5* taneously catalytically active. This would Glucose-6-phosphatase Fed 9.9 63-1 an a Starved 15-1* 90.3* then provide example of substrate Sepsis 9 0 69-6 cycle (the glucose/glucose-6-phosphate cycle), Phosphofructokinase Fed 12-3 72-7 Starved 9.9 57-3* which occurs when a non-equilibrium reaction Septic 8-0* 57-7* proceeds in both the forward and reverse direc- Lactate dehydrogenase Fed 102 580 Starved 95-6 560 tions ofthe pathway simultaneously. This cycle Septic 81-7 585 is considered to play an important part in the Pyruvate kinase Fed 23-2 122-7 Starved 18-1* 101-3 control of glucose utilisation by the liver.2 Septic 13-5* 96-2 The cycle would demand a considerable Pyruvate dehydrogenase Fed 0-28 1-62 in Starved 0.19* 1 18* expenditure ofenergy, which might account, Sepsis 0.39* 2-81* part, for some of the thermic effect of food. Pyruvate carboxylase Fed 0-41 2-2 a Starved 0-43 2-6 The intriguing question is, however, why Sepsis 0-26 2-0 substrate cycle should restrict glucose utilisa- tion in this tissue. The answer is not known. Results are presented as means. The temperature of assay was 25°C except for glucose-6- The of a in the phosphatase (30°C). The statistical difference (Student's t test) of the difference in enzyme presence cycle enterocytes activities in the intestine of septic animals - with that of fed animals is shown by * (p<005). would account for the findings of several Quantitative aspects ofglucose and glutamine metabolism by intestinal cells S15

TABLE III Maximal capacityfor glucose utilisation in TABLE V Rates of utilisation ofglutamine by varinous in various rat tissues vitro cell preparations of the rat and mouse Hexokinase activity Type ofcell Rate of utilisation (Qimol/hig dry wt) Gut: first published as 10.1136/gut.35.1_Suppl.S13 on 1 January 1994. Downloaded from Tissues pmollmin/g p.mol/min/rat Rat enterocytes 220 Rat colonocytes 730 Intestine 11-0 33 Rat lymphocytes 331 Brain 8 7 21-2 Rat thymocytes 160 Kidney 2-1 4-2 Murine macrophages 186

workers that only a very small proportion of the activity increases. In addition, in starvation, glucose taken up from the lumen of the the concentration of glucose-6-phosphate in intestine is metabolised despite the very high the intestine is considerably decreased. This hexokinase activity in these cells.7 8 Similar low suggests that the control of glycolysis occurs at rates of glucose utilisation have been shown for an earlier step than that catalysed by phospho- humans.9 fructokinase, that is, at the step of glucose These findings also suggest that glucose is phosphorylation by the glucose/glucose-6- not normally an important fuel for energy phosphate cycle. The existence of such a cycle generation in the intestine. The fact that in the small intestine has not, however, been isolated incubated enterocytes or the perfused investigated in any detail. intestine, however, can take up and use glucose at a high rate does suggest that, under some conditions, glucose may become an important Glutamine utilisation fuel. Unfortunately, the way in which the The first reaction seen in glutamine degrada- activities of hexokinase and glucose 6-phos- tion is catalysed by a phosphate dependent glu- phatase are controlled in the intestine is not taminase (a mitochondrial enzyme hereafter known. referred to as glutaminase). Glutaminase's activity in the intestine is higher than that in most other tissues (Table IV). Hypothesis for the control of glucose Glutamine could also be degraded to gluta- metabolism during starvation mate by a reaction catalysed by a glutamine Starvation decreases the glucose metabolism aminotransferase, but this enzyme has much rate in tissues, in several ofwhich an important lower activity than that of glutaminase in the mechanism for the regulation of glucose utili- intestine, suggesting that most glutamine is sation is the glucose/fatty acids/ketone body converted by glutaminase. cycle.2 Fatty acid or ketone body oxidation, or The activity of glutamate dehydrogenase in both, raises the intracellular concentration of intestinal mucosa is low compared with that citrate and acetyl CoA, and this inhibits phos- found in the kidney; this, together with the phofructokinase and pyruvate dehydrogenase, comparatively high aspartate and alanine http://gut.bmj.com/ respectively. This is, however, probably not the aminotransferase activities, suggests that gluta- mechanism in the small intestine because the mate is metabolised by an aminotransferase rate of glucose utilisation by the vascularly rather than by glutamate dehydrogenase. perfused intestine in vitro is unchanged by the Indeed, the important end products of gluta- utilisation of ketone bodies.10 Furthermore, mine metabolism in all intestinal preparations the concentrations of ATP, citrate, and are alanine and CO2. Such high decarboxyla- glucose-6-phosphate in the intestinal mucosa tion rates of glutamine may occur if 2- on September 23, 2021 by guest. Protected copyright. are not increased during starvation, so these oxoglutarate is first converted to pyruvate, if metabolites probably control the glycolytic pyruvate is then converted to acetyl CoA, and flux. Hence, an alternative mechanism for the subsequently decarboxylated by the reactions regulation of glucose metabolism in the of the Krebs' cycle. intestine has to be found. Hexokinase and The intestinal mucosa contains various cell glucose-6-phosphatase are present in the types (for example, enterocytes and immune intestine of fed rats at similar activities, but cells). As immune cells (lymphocytes and in the starved state, hexokinase activity macrophages) are known to use glutamine at a decreases, whereas glucose-6-phosphatase high rate (see below), how do enterocytes, lymphocytes, and macrophages contribute to TABLE IV Effects ofsepsis on glutaminase activity in rat tissue the high rate of glutamine uptake by the small intestine? Unfortunately, the exact proportion Enzyme activities of lymphocytes and enterocytes in the intesti- Tissue Nutritional state ,Amol/min-l/gfresh wt nmol/min -lmg protein nal mucosa is not yet firmly established. As much as 25% of the mucosal mass may be Intestinal mucosa Fed 12 9 81-0 Starved 9.4* 59.3* lymphoid tissue, and 80% of the intestinal Septic 7-1* 52.1* Mesenteric lymph node Fed 3-3 18-9 epithelium may be comprised of enterocytes. Starved 4-4 23-8* Hence, it can be assumed that enterocytes con- Septic 5 1* 28-2* the mass Kidney Fed 27-4 172 tribute most to the 750/o of mucosal Starved 25-7 163 that is non-lymphoid. The rate of glutamine Septic 29-5 188 Heart Fed 0 95 4-7 utilisation by isolated lymphocytes is similar to Starved 0 7 3-6* that of enterocytes (Table V).5 Hence, the Septic 0-69* 3-2* whole intestinal glutamine metabolism could be attributed to that of the enterocytes Results are presented as means. The temperature of assay was 37°C. The statistical difference plus (Student's t test) between septic, starved, and fed rats is shown by * (p<005). lymphocytes. S16 Newsholme, Canie

An important difference between metab- response to sepsis, as has been shown in olism of glutamine by isolated enterocytes and incubated enterocytes that decreased their rate that by lymphocytes is the nitrogen containing ofglucose utilisation by 26%. The proportion of

end products of metabolism. Isolated prepara- glucose that is metabolised to lactate, pyruvate, Gut: first published as 10.1136/gut.35.1_Suppl.S13 on 1 January 1994. Downloaded from tions of enterocytes produce alanine, whereas and alanine is, however, unchanged in sepsis. lymphocytes produce aspartate. This suggests Although rates of glucose utilisation by that the enterocytes rather than immune cells isolated enterocytes are decreased in response are responsible for much of the glutamine to sepsis, rates of pyruvate decarboxylation are metabolism in isolated preparations of the increased. Hence, 'efficiency' of energy intestinal mucosa. generation from glucose or glutamine may Glutamine utilisation in lymphocytes (and improve under septic conditions. macrophages) provides not only energy but also metabolic intermediates necessary for biosynthetic pathways. The very high rate GLUTAMINE METABOLISM of glutaminolysis in lymphocytes (and Intestinal glutaminase activity decreases in macrophages), however, may provide optimal sepsis, whereas that in lymphocytes increases conditions for controlling the rate of biosyn- (Table IV). It is possible that decreased gluta- thetic processes at crucial times. This permits minase intestinal activity and glutamine uptake increased macromolecular synthesis at precise together ensure adequate glutamine supply for times during the cell cycle, such as when the immune cells, whose number and activity lymphocytes respond to an immunological would be increased under sepsis. Under these challenge.1' 12 Enterocytes isolated in this conditions, however, hexokinase activity in the work are not believed to be dividing, nor intestine also decreased, but the rate of 14CO2 secreting any products (as macrophages would production from [14CU]-glucose was increased be). The question arises, therefore, why compared with that in the fed state. Hence, the enterocytes use glutamine at such a high rate. decreased rate of glutamine utilisation could Perhaps it provides energy, which might be instead 'allow' more glucose to be oxidised to required, for example, for active transport or provide energy. These increased rates of 14CO2 for maintaining the integrity of the intestinal production from the radiolabelled substrates wall. Thus, intestinal utilisation of glutamine (glucose and glutamine) in sepsis were paral- may 'spare' glucose metabolism. Glucose leled by increased pyruvate dehydrogenase and absorbed luminally by the enterocytes, then, is oxoglutarate dehydrogenase activities. The not metabolised to any large extent, but is mechanism by which these changes occur is released into the blood to maintain the unknown but it suggests that investigation of blood glucose concentration available to the control of pyruvate metabolism in entero- other tissues (for example brain). 13 14 In this cyte mitochondria may be a fruitful line of way, the intestine could be considered to research.

play an important part in regulation of the Rates of glutamine utilisation and rates of http://gut.bmj.com/ blood glucose concentration. This could be glutamate and alanine production by isolated of considerable importance in a number of enterocytes are decreased in response to conditions, but especially in diabetes mellitus. sepsis.'6 These changes, together with the This may be an important area for new decrease in glutaminase activity, support the research. view that sepsis may decrease the rate of gluta- mine utilisation by the small intestine in vivo.

This change in the rate of glutamine utilisation on September 23, 2021 by guest. Protected copyright. Effects of sepsis in sepsis is in contrast with that seen in Intestinal fuel utilisation may be changed by thermally injured rats which increase the rate starvation, injury, surgery, and sepsis. Sepsis of glutamine utilisation.17 This increased rate (by endotoxin injection) affects both glucose may subsequently decrease glucose utilisation and glutamine metabolism.'5 locally to make more glucose available for repair of the burn wound, which requires high rates of glucose utilisation. GLUCOSE METABOLISM Changes in the rate of glutamine utilisation The maximal in vitro activities of some by the small intestine in sepsis may permit enzymes participating in glucose metabolism more glutamine to be available to the cells were measured in intestinal mucosal extracts. participating in the immune process, because Sepsis decreased the maximal activities of all the glutamine requirement of those cells will the glycolytic enzymes; this reduction of probably increase during sepsis. Fluctuations activities was quantitatively similar to that of in intestinal glutamine utilisation may provide starvation. One important difference is that, in precise control of some biosynthetic pathways contrast with starvation, sepsis had no effect on in immune cells." 12 Sepsis is also character- the maximal activity of glucose-6-phosphatase ised by a decreased rate of glucose utilisation (Table I). This is of considerable interest in by the intestine, and as both high rates of sepsis, as the activity of pyruvate dehydro- glutaminolysis and glycolysis are required by genase was increased by 73% compared with cells of the immune system, blood glucose may the normal (fed) state, and by 138% compared also be directed preferentially to these cells. with the starved state. The question arises as to why the intestine, The decrease in the maximal activities of the paradoxically, has a reduced ability to utilise glycolytic enzymes suggests a reduced ability glutamine during sepsis despite the rich of the small intestine to utilise glucose in lymphoid population of the mucosa. It is Quantitative aspects ofglucose and glutamine metabolism by intestinal cells S17

possible that overall glutamine utilisation in the portal vein and not metabolized to lactate. Biochem?i 7 1988; 254: 931-4. the intestine during sepsis is the net result of a 9 Bjorkman 0, Eriksson LS, Nyberg B, Wahren J. Gut decreased rate of glutamine utilisation by exchange of glucose and lactate in basal state and after oral glucose ingestion in post-operative patients. Diabetes enterocytes, which exceeds the increased rate 1990; 39: 747-51. Gut: first published as 10.1136/gut.35.1_Suppl.S13 on 1 January 1994. Downloaded from of utilisation by the immune cells of the 10 Hanson PJ, Parsons DS. Factors affecting the utilization of ketone bodies and other substrates by rat jejunum: effects intestine. of fasting and of diabetes. I Physiol (London) 1976; 278: 55. 11 Newsholme EA, Crabtree B, Ardawi MSM. The role of 1 Newsholme EA, Crabtree B, Newsholme P. Use of enzyme high rates of glycolysis and glutamine utilization in activities as indices of maximum rates of fuel utilization. rapidly-dividing cells. Biosci Rep 1985; 4: 393-400. Ciba Found Symlp 1980; 73: 245-58. 12 Szondy Z, Newsholme EA. The effect of glutamine concen- 2 Newsholme EA, Leech AR. Biochemtistry for the mledical tration on the activity of carbamoyl-phosphate synthase II sciences New York: John Wiley, 1983. and on the incorporation of [3H]-thymidine into DNA in 3 Ardawi MSM, Newsholme EA. Metabolism in lymphocytes rat mesenteric lymphocytes stimulated by phytohaemag- and its importance to the immune response. Essays glutin. Biochenj 1989; 261: 979-83. Biochein 1985; 21: 1-44. 13 Fernandez Lopez JA, Casado J, Argiles JM, Alemany M. In the rat intestinal lymph carries a significant amount of 4 Ardawi MSM, Newsholme EA. Maximum activities of ingested glucose into the bloodstream. Arch Int Physiol some enzymes of glycolysis, the tricarboxylic acid cycle Biochimti Biophys 1992; 100: 231-6. and glutamine utilization pathways in lymphocytes of the 14 Fernandez Lopez JA, Casado J, Argiles JM, Alemany M. rat. BiochemnJ_ 1982; 208: 743-9. Intestinal handling of a glucose gavage by the rat. 5 Ardawi MSM, Newsholme EA. Glutamine metabolism in Molecular Cellular Biochemtz 1992; 113: 43-53. lymphocytes of the rat. BiochemnJ 1983; 212: 835-42. 15 Parry-Billings M. Studies opn glutampinle mizetabolismn MPnuscle. 6 Newsholme P, Newsholme EA. Rates of utilization of D Phil [Thesis], Oxford: Oxford University, 1989. glucose, glutamine and oleate and formation of end- 16 Carrie A-L. Studies on glucose anid glutamtine mtietabolismn ini products by mouse peritoneal macrophages in culture. cells of the smtiall intestinie [Thesis]. Oxford: Oxford Biochemn1 989; 261: 211-8. University, 1989. 7 Windmueller HG, Spaeth AE. Respiratory fuels and 17 Ardawi MSM, Newsholme EA. Maximal activities of gluta- nitrogen metabolism in vivo in small intestine rats. J Biol minase and some enzymes of glycolysis and ketone body Chein 1980; 255: 107-22. utilization and rates of utilization of glutamine glucose 8 Rich-Denson C, Kimure RE. Evidence in vivo that most of and ketone bodies by intestinal mucosa after burn injury. the intraluminally absorbed glucose is absorbed intact into Bunts 1987; 13: 438-44. http://gut.bmj.com/ on September 23, 2021 by guest. Protected copyright.