• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of biochemjBJ Latest papers and much more!
Biochem J. Jan 15, 2007; 401(Pt 2): 465–473.
Published online Dec 21, 2006. Prepublished online Sep 27, 2006. doi:  10.1042/BJ20061148
PMCID: PMC1820798

Glutamine gluconeogenesis in the small intestine of 72 h-fasted adult rats is undetectable


Recent reports have indicated that 48–72 h of fasting, Type 1 diabetes and high-protein feeding induce gluconeogenesis in the small intestine of adult rats in vivo. Since this would (i) represent a dramatic revision of the prevailing view that only the liver and the kidneys are gluconeogenic and (ii) have major consequences in the metabolism, nutrition and diabetes fields, we have thoroughly re-examined this question in the situation reported to induce the highest rate of gluconeogenesis. For this, metabolically viable small intestinal segments from 72 h-fasted adult rats were incubated with [3-13C]glutamine as substrate. After incubation, substrate utilization and product accumulation were measured by enzymatic and NMR spectroscopic methods. Although the segments utilized [13C]glutamine at high rates and accumulated 13C-labelled products linearly for 30 min in vitro, no substantial glucose synthesis could be detected. This was not due to the re-utilization of [13C]glucose initially synthesized from [13C]glutamine. Arteriovenous metabolite concentration difference measurements across the portal vein-drained viscera of 72 h-fasted Wistar and Sprague–Dawley rats clearly indicated that glutamine, the main if not the only gluconeogenic precursor taken up, could not give rise to detectable glucose production in vivo. Therefore we challenge the view that the small intestine of the adult rat is a gluconeogenic organ.

Keywords: gluconeogenesis, glucose synthesis, glutamine metabolism, ketone body, NMR spectroscopy, small intestine
Abbreviations: PEPCK, phosphoenolpyruvate carboxykinase


The contribution of the liver to glucose production in normal fasted humans and animals is well established; under this situation, the liver produces glucose as a result of both gluconeogenesis mainly from lactate, alanine, glutamine and glycerol and glycogenolysis when glycogen stores are available [1,2]. Various studies performed in humans and, to a lesser extent, in rats have suggested that, besides the liver, the kidneys may contribute to systemic gluconeogenesis; indeed, it has been reported that renal gluconeogenesis may represent 5–45% of systemic gluconeogenesis in post-absorptive or fasted humans (see [3] for a recent review).

Combining isotopic techniques with or without arteriovenous concentration difference measurements or organ vessel ligature, Mithieux and co-workers [47] recently published a series of papers assessing the contribution of the small intestine of adult rats to systemic endogenous glucose production under various experimental conditions. In agreement with the findings of Watford et al. [8] and of Baverel and Lund [9] with isolated enterocytes from fed rats, they did not observe any intestinal gluconeogenesis in vivo in control rats fasted for 5–6 or 24 h [4,5], although they more recently found the reverse in 24 h-fasted animals [6,7]. In addition, no intestinal glucose synthesis was observed in post-absorptive or 24 h-fasted rats in vivo, especially from [14C]glutamine [4,5]. By contrast, they concluded from their data that the small intestine of the rat accounted for 21, 35, 19 and 19% of systemic gluconeogenesis after 48 or 72 h of fasting, after 3 days of Type 1 diabetes and 2–15 days of a high-protein diet respectively [47]. These authors also found that, in the rat small intestine, an increase in the expression and activity of PEPCK (phosphoenolpyruvate carboxykinase) and glucose-6-phosphatase, two key gluconeogenic enzymes, was consistent with the in vivo induction of gluconeogenesis [5,10,11]. Furthermore, they claimed that the satiety induced by a high-protein diet was linked to the increase in intestinal gluconeogenesis caused by such a diet [6]. Thus, in the light of the latter studies, intestinal gluconeogenesis would become a central process controlling not only glucose homoeostasis but also the whole energy metabolism of the body.

The observations and conclusions of Mithieux and co-workers [4,7] are intriguing because, unlike the liver and the kidney, the small intestine of the rat has until recently never been considered to be a site of intense gluconeogenic capacity. To our knowledge, only Windmueller and Spaeth [12,13] previously observed in vivo that the small intestine of the adult rat has a very low capacity to synthesize radioactive glucose from radioactive aspartate, glutamine and lactate. In vitro preparations of the small intestine from suckling but not weaned rats have also been reported to synthesize some glucose from lactate [14].

Given (i) the potential importance of the small intestine as a gluconeogenic organ and as the source of major glucose homoeostasis disturbances, (ii) the absence of glucose synthesis from glutamine in isolated enterocytes from fed rats shown by Watford et al. [8] and Baverel and Lund [9], (iii) that the conclusions drawn by Mithieux and co-workers [4,7] should be taken with caution, as pointed out by the authors themselves because of inaccuracies inherent to the numerous experimental steps needed to determine the gluconeogenic rates they reported [15], and (iv) the recent review by Watford [16] casting serious doubt on the existence of intestinal gluconeogenesis, we have decided to thoroughly re-examine this question both in vitro and in vivo. For this, we have isolated metabolically viable segments of the small intestine obtained from rats subjected to a 72 h fast, an experimental condition reported so far to be the most potent inducer of intestinal gluconeogenesis [5]. These segments, made of the entire intestinal wall to mimic in vitro the structure of the small intestine in vivo, were incubated with 13C-labelled glutamine. Taking advantage of the 13C NMR spectroscopy and enzymatic approaches that enabled us to unravel the complexity of the renal metabolism of various substrates [1722], we measured glutamine utilization and product accumulation. Despite high rates of glutamine metabolism, substantial glucose synthesis could not be detected. Furthermore, arteriovenous concentration difference measurements were performed across the intestine not only in 72 h-fasted Wistar but also in 72 h-fasted Sprague–Dawley rats because the latter rat strain was used by Mithieux and co-workers [47]. The results obtained revealed that the small percentage of glutamine taken up by the intact intestine cannot explain the large glucose production reported by Mithieux and co-workers in their rats under the same experimental conditions. A preliminary account of part of the present results has been published in abstract form [23].



Glucosamine and glutaminase (grade V) were from Sigma (St Quentin-Fallavier, France). Other enzymes and coenzymes were purchased from Boehringer Mannheim (Meylan, France). L-[3-13C]Glutamine and D-[2-13C]glucose were obtained from Eurisotop (Saclay, France) and had a 99% isotopic abundance.


All experiments were approved by the Institutional Animal Care and Use Committee of the Lyon 1 University. Male Wistar or Sprague–Dawley rats weighing between 250 and 300 g were obtained from Charles River (Saint-Germain-sur-l'Arbresle, France). They were acclimated to our animal house for 1 week, kept on a 12 h light/12 h dark cycle and fed a standard diet (U.A.R., Villemoisson-sur-Orge, France). Then, they were fasted for 72 h before the experiments but had free access to water.

Preparation and incubation of intestinal segments from 72 h-fasted Wistar rats

Each rat was anaesthetized with intraperitoneal Nembutal (50 mg·g body weight−1). The abdomen was opened and the small intestine was exteriorized, quickly excised and stripped away from the mesentery. The lumen of the small intestine was flushed out with twice 60 ml of ice-cold oxygenated Krebs–Henseleit medium [23a]. All the subsequent steps of preparation were performed in cold oxygenated Krebs–Henseleit buffer. Then, rings of 2–4 mm width were cut with scissors from the entire length of the small intestine. The rings were pooled and randomized in a flask and then separated from the buffer on a sieve, gathered on a cold glass surface and divided into five equivalent batches of rings that were immediately incubated. Two batches of rings were incubated per incubation time (20 and 30 min) in 250 ml Erlenmeyer flasks containing an atmosphere of O2/CO2 (19:1) and 20 ml of Krebs–Henseleit buffer without or with 5mM L-[3-13C]glutamine, 1 or 5 mM unlabelled or D-[2-13C]glucose. Glucosamine (40 mM), an inhibitor of hexokinase [24], was also used in certain experiments. The flasks were incubated at 37 °C in a shaking water bath. Incubation was terminated by the addition of HClO4 [final concentration 5% (v/v)] and rapid homogenization of tissue plus medium. A batch of rings was taken to determine the initial amounts of added and endogenous substrates by adding HClO4 before the rings. After centrifugation (3000 g for 5 min), the denatured rings were used for measurement of their protein content; the supernatant was neutralized with a mixture of 20% (w/v) KOH and 1% (v/v) H3PO4 and used for metabolite determination and NMR spectroscopy.

Measurement of arteriovenous concentration differences

Anaesthetized, 72 h-fasted Wistar and Sprague–Dawley rats (250–300 g) were placed on a heated table to maintain their body temperature at 37–38 °C. Then, a catheter was inserted into the carotid artery. After laparotomy, the gut was gently moved on the left side and covered with gauze moistened with warm saline. Hepatic portal blood was collected in the following way: a needle (gauge 23) was bent, mounted on a pipette holder, connected to a syringe by a length of polyethylene tubing and introduced into the portal vein by puncture. For blood withdrawing, the position of the needle tip inside the hepatic portal vein was continuously controlled thanks to a binocular lens with a magnification of ×8. Blood samples (1.5 ml each) were aspirated slowly (over 60 s) and simultaneously from the carotid artery and the hepatic portal vein. This procedure was never accompanied by any bleeding. Immediately after blood collection, 1.3 ml of blood sample was deproteinized with 0.5 ml of 20% (v/v) chilled HClO4. After centrifugation, the supernatants were neutralized with a mixture of KOH and phosphoric acid as described above and used for metabolite analysis.

Analytical methods

ATP and protein content

The intestinal tissue (rings) ATP concentration was quantified by using the method of Lamprecht and Trautschold [25]. Pellets were solubilized in 0.5 M NaOH for protein determination and total protein was determined as described previously [26].

Metabolite assays

Glutamine, glutamate, alanine, lactate, pyruvate, aspartate, glucose, ammonia, glycerol, D-β-hydroxybutyrate, acetoacetate and glycogen were determined according to the method of Passonneau and Lowry [27].

13C NMR techniques

Data were recorded as indicated previously [17,18,22] at 125.75 MHz on a Bruker AM-500 WB spectrometer using a 5 mm broadband probe thermostatically maintained at 8±0.5 °C, except that the number of scans was 420. Chemical shifts were expressed as p.p.m. (parts per million) relative to tetramethylsilane. Assignments were made by comparing the chemical shifts obtained with those given in the literature [28,29]. Given that glutamine gluconeogenesis from [3-13C]glutamine leads to the equal labelling of the C-1, C-2, C-5 and C-6 of glucose [21], and taking a signal/noise ratio equal to 1.5 for the heights of the corresponding β resonances of glucose carbons, our calculations indicate that, under our experimental conditions, approx. 0.7 μmol of the C-3 of [3-13C]glutamine had to be converted into glucose to enable us to detect glutamine gluconeogenesis.

Calculations and statistical analysis

Net substrate utilization and product formation were calculated as the difference between the total flask contents (tissue plus medium) at the start (zero-time flasks) and after the period of incubation. The net metabolic rates, reported as means±S.E.M., are expressed in μmol of substance removed or produced per unit time (20 or 30 min) per g of intestinal tissue protein.

With [3-13C]glutamine as substrate, the transfer of the C-3 of glutamine to a given position in a given metabolite was calculated by using the formula described previously [17,18,22].

The blood metabolite concentrations found in the carotid artery and the hepatic portal vein were compared using the paired Student's t test. P<0.05 was considered to be statistically significant.


Glutamine metabolism in intestinal segments from 72 h-fasted rats

Since the viability of isolated rat enterocytes is limited with time [8], our segments were incubated for 20 and 30 min, but, for accuracy purposes, these short incubation times were compensated by the incubation of large amounts of intestinal tissue in a large volume (20 ml) of incubation medium. At zero time, the ATP content of intestinal segments was 2.99±0.71 μmol·g of protein−1 (n=3). In intestinal segments incubated with glutamine, the ATP content was 2.90±0.49 and 2.90±0.39 μmol·g of protein−1 after 20 and 30 min of incubation respectively (n=3).

Table 1 shows the time course of glutamine utilization and product formation measured by enzymatic methods when the segments were incubated with 5 mM [3-13C]glutamine as substrate. Glutamine utilization occurred at high rates and in a linear fashion with time. The rate of glutamate accumulation was approximately constant, whereas those of alanine, lactate and ammonia tended to increase with time. Small amounts of aspartate, urea and pyruvate accumulated and negligible amounts of glucose were synthesized. In the absence of glutamine, substantial amounts of alanine, lactate and ammonia, and very small amounts of glucose, accumulated from endogenous substrates (Table 1).

Table 1
Metabolism of 5 mM L-[3-13C]glutamine in small intestinal segments from 72 h-fasted rats

Figure 1(A) shows a representative 13C NMR spectrum of perchloric extracts obtained after 30 min of incubation of intestinal segments from a 72 h-fasted rat with 5 mM [3-13C]glutamine. This spectrum, in which all the most significant resonances could be identified, shows the main non-volatile carbon products of glutamine metabolism. Not only glutamate but also alanine and lactate were the most important labelled products found. The fact that the C-2 of glutamate became labelled during incubation indicates that glutamate was re-synthesized after passage of the C-3 of glutamine through the tricarboxylic acid cycle. Small amounts of the C-3 of glutamine were also incorporated into aspartate, proline and ornithine. No labelling of glucose carbons was observed, as shown in the upper left panel of Figure 1(A) showing the baseline of the spectrum where the glucose resonances appear (in the 100−60 p.p.m. region).

Figure 1
13C NMR spectra (125.17 MHz) of neutralized HClO4 extracts obtained from small intestinal segments (from 72 h-fasted Wistar rats) incubated for 30 min with [3-13C]glutamine in the absence (A) and the presence (B) of glucosamine ...

From the spectra obtained after 20 and 30 min of incubation of segments from 72 h-fasted rats, we calculated the amounts of [3-13C]glutamine removed and of labelled products accumulated after correction for the 13C natural abundance. As shown in Table 2, the removal of the labelled glutamine was also approximately linear with time and this removal was not statistically different with that measured enzymatically (see Table 1). This means that, in agreement with the very low activity of glutamine synthetase in the small intestine [30], there was no substantial glutamine synthesis in our experiments. The labelling of glutamate, alanine and lactate was lower than the amounts of these products found to accumulate (see Table 1); this was especially true for alanine and to a lesser extent for glutamate and lactate. The fact that the labelling of aspartate was higher than the net aspartate accumulation presented in Table 1 simply reflects the fact that the unlabelled aspartate brought by the segments at zero time was metabolized and replaced by newly synthesized labelled aspartate. Table 2 also shows that the labelled C-2 of aspartate, alanine and lactate was virtually equal to the labelled C-3 of these compounds; this is in agreement with the view that the C-3 of glutamine converted into the C-3 of glutamate by glutaminase and then into the C-3 of 2-oxoglutarate by alanine, aspartate and ornithine aminotransferases, was further metabolized in the tricarboxylic acid cycle via succinate and fumarate, two symmetrical molecules. The passage through these symmetrical molecules is also consistent with the labelling of the C-2 of glutamate which involves one complete turn of the tricarboxylic acid cycle. The labelling of the C-1 of lactate and alanine is also consistent with the recycling of the C-3 of glutamine through the tricarboxylic acid cycle to yield the C-1 of malate and then the C-1 of pyruvate by malic enzyme, or the C-1 of oxaloacetate and then the C-1 of pyruvate by oxaloacetate decarboxylase or PEPCK plus pyruvate kinase [8]. In agreement with the results of Windmueller and Spaeth who used [14C]glutamine as substrate [13,31,32], some 13C-labelled ornithine, citrulline and proline were also formed from 13C-labelled glutamine. It is noteworthy that the sum of the non-volatile labelled products of [13C]glutamine metabolism increased linearly with time. Neglecting the possibility that a small fraction of the [13C]glutamine removed was incorporated into tissue materials [13,31,32], we calculated the 13CO2 produced from [3-13C]glutamine to estimate the complete oxidation of glutamine carbon skeletons, as explained in a previous publication [21]. After both 20 and 30 min of incubation, approximately half of the C-3 of glutamine used was converted into CO2 (59.4±8.9 and 81.2±11.2 μmol·g of protein−1·incubation time−1 respectively). The corresponding proportions of glutamine carbon converted into CO2 are close to those found in the rat small intestine in vivo [13,31,32], indicating that all steps of the tricarboxylic acid cycle were functioning well in our segments in vitro.

Table 2
Metabolism of 5 mM L-[3-13C]glutamine in small intestinal segments from 72 h-fasted rats

Evidence that gluconeogenesis from glutamine in small intestinal segments from 72 h-fasted Wistar rats was not masked by concomitant glucose re-utilization

Given the high capacity of the small intestine to metabolize glucose in vitro [8,33,34], we used two experimental approaches to test the possibility that the [13C]glucose synthesized from part of the [3-13C]glutamine metabolized by our intestinal fragments was reutilized via the glycolytic pathway. As done in a recent study performed with precision-cut human kidney slices [22], we hypothesized that the [3-13C]glutamine not accounted for by the non-volatile 13C-labelled products found to accumulate might have been initially converted into [13C]glucose and then re-utilized and oxidized into 13CO2 by our intestinal segments. It can be calculated that, if all this synthesized [13C]glucose had accumulated in the incubation medium, its concentration would have been approx. 1 mM.

Effect of glucosamine on the metabolism of 5 mM [3-13C]glutamine

First, we tested the efficacy of glucosamine, an inhibitor of hexokinase [24], on glucose utilization by small intestinal segments from 72 h-fasted Wistar rats. With 1 mM D-[2-13C]glucose as substrate, addition of 40 mM glucosamine induced a mean 34% inhibition of the removal of [2-13C]glucose, measured by NMR spectroscopy; the corresponding values were 55.7±1.5 and 36.6±1.8 μmol·g of protein−1·30 min−1 in the absence and the presence of glucosamine respectively (n=3; P<0.05).

Figure 1(B) shows a 13C NMR spectrum obtained in the presence of 40 mM glucosamine in one representative experiment out of two performed in duplicate in which small intestinal segments from 72 h-fasted Wistar rats were incubated for 30 min with 5 mM [3-13C]glutamine in the absence and the presence of 40 mM glucosamine. This spectrum, which shows the resonances corresponding to the C-3 of glutamine and those of the 13C-labelled products of [13C]glutamine (as in Figure 1A) and the resonances of the glucosamine carbons, clearly demonstrates that there were no glucose carbon resonances. In this respect, it should be pointed out that the resonances of the C-1 of glucose, which are well separated from those of the C-1 of glucosamine, were absent.

Effect of 5 mM glucose on the metabolism of 5 mM [3-13C]glutamine in small intestinal segments from 72 h-fasted Wistar rats

With the hope that large amounts of glucose added at zero time would (i) dilute the 13C-labelled glucose possibly synthesized from [3-13C]glutamine and (ii) prevent its complete re-utilization, four additional experiments were performed with small intestinal segments (265.8±8.0 mg of protein·flask−1) from 72 h-fasted Wistar rats incubated for 30 min with 5 mM [3-13C]glutamine in the absence and the presence of 5 mM unlabelled D-glucose. Under the latter conditions, glutamine utilization was 136.5±8.9 and 142.4±10.1 μmol·g of protein−1·30 min−1 respectively. In the presence of 5 mM glucose, glucose utilization was 148.8±15 μmol·g of protein−1·30 min−1. Figure 2 presents a 13C NMR spectrum obtained in the presence of 5 mM [3-13C]glutamine+5 mM unlabelled D-glucose in one representative experiment. This spectrum, which shows the resonance of the C-3 of glutamine and those of the 13C-labelled products of [13C]glutamine metabolism (as in Figure 1), clearly demonstrates the presence of glucose carbon resonances. However, as suggested by the fact that all the carbon resonances (the sums of the anomers α and β) had virtually the same intensities (peak heights) and peak areas, these resonances corresponded exclusively to the natural abundance of the glucose carbons added at zero time and not metabolized by the intestinal segments after 30 min of incubation. Therefore no synthesis of 13C-labelled glucose carbons occurred from [3-13C]glutamine in these experiments.

Figure 2
13C NMR spectrum (125.17 MHz) of neutralized HClO4 extracts obtained from small intestinal segments (from 72 h-fasted Wistar rats) incubated with [3-13C]glutamine in the presence of unlabelled glucose

Metabolite concentrations in arterial and portal vein blood of 72 h-fasted Wistar and Sprague–Dawley rats

Table 3 shows the mean concentrations of all the metabolites measured in the arterial and portal vein blood of 72 h-fasted rats. In both strains, there was a net uptake of glucose; this markedly contrasts with the absence of net glucose uptake or release observed by Mithieux et al. [5] in their 72 h-fasted Sprague–Dawley rats. In agreement with the classical results obtained by numerous authors in fed rats or in rats fasted for up to 48 h [13,31,3538], we observed a 30% uptake of arterial glutamine by the intestine in our 72 h-fasted Wistar rats, whereas the corresponding value was only 18% in our 72 h-fasted Sprague–Dawley rats.

Table 3
Arterial and portal vein blood concentrations of metabolites in 72 h-fasted Wistar and Sprague–Dawley rats

A statistically significant release of glutamate occurred only in Sprague–Dawley rats and this release represented 60% of the glutamine taken up in these animals. A release of alanine, which was of the same order of magnitude in Wistar and Sprague–Dawley rats, was observed.

In agreement with the observation of Windmueller and Spaeth [13] in their overnight-fasted Osborne–Mendel rats when arterial lactate was higher than 1.2–1.6 mM, the intestine of our Wistar rats took up a large fraction (37%) of circulating lactate; by contrast, the intestine of our Sprague–Dawley rats neither took up nor released lactate in a statistically significant manner. A very small uptake of glycerol occurred only in Sprague–Dawley rats. D-β-Hydroxybutyrate, whose arterial concentration was much higher in Wistar than in Sprague–Dawley rats, was taken up only by the intestine of Wistar rats and part of this D-β-hydroxybutyrate was released as acetoacetate which was neither taken up nor released by the intestine of Sprague–Dawley rats.


Although we used Wistar rats after 72 h of fasting, an experimental condition reported in vivo to induce a large intestinal gluconeogenesis [5], the results of the present study unequivocally demonstrate that, when incubated in vitro, the small intestine of our animals did not synthesize glucose from glutamine, the main if not the exclusive gluconeogenic precursor after 48 h [4] and presumably 72 h of fasting. In that respect, it resembles the isolated perfused and intact functioning small intestine of overnight-fasted rats [31,32] and the enterocytes isolated from fed rats in which no glutamine gluconeogenesis was reported [8,9]. It should be mentioned here that, within the intestinal wall, the enterocytes are the only site of glutamine metabolism [31]. Moreover, our in vivo results do not corroborate those of Mithieux et al. [5] who concluded that the small intestine of 72 h-fasted Sprague–Dawley rats is responsible for at least one-third (35%) of systemic gluconeogenesis.

Metabolic viability of our small intestinal fragments and absence of glucose synthesis

Several lines of evidence indicate that our intestinal segments, which consisted of the entire small intestinal wall, were metabolically viable. Taking a dry weight/fresh weight ratio equal to 0.23 [31] and assuming that 90% of the dry weight consisted of protein, it can be calculated that the small intestine preparation used by Windmueller and Spaeth [32] metabolized the luminally added glutamine (6 mM) at a rate of 27 μmol·g of protein−1·30 min−1. Taking the glutamine removed in our experiments after 30 min of incubation (Table 1), it can also be calculated that our small intestinal segments, a maximum of 40% of which consisted of enterocytes [39,40], metabolized 5 mM glutamine at a much higher rate (586 μmol·g of protein−1·30 min−1). The enterocytes isolated by Watford et al. [8] and by Baverel and Lund [9] metabolized 5 mM glutamine at a rate of 367 and 300 μmol·g of protein−1·30 min−1 respectively. Thus, although our measurements and calculations reveal that the ATP level of our segments was only about one-half that found in vivo [41], and although the enterocytes represented only approx. 40% of our intestinal segments, they were metabolically very active and metabolized glutamine avidly and in a virtually linear manner with time (Tables 1 and and2).2). It is also important to note that, like the glutamine taken up from the lumen and from the circulating blood in the in vivo preparations used by Windmueller and Spaeth [13,31], about half of our glutamine metabolized was completely oxidized (see the Results section). This means that the mitochondria of our intestinal segments functioned satisfactorily. That the entire tricarboxylic acid cycle was operating in a satisfactory manner is also indicated by the conversion of the C-3 of glutamine not only into the C-2 and C-3 of alanine and lactate but also into the C-2 of glutamate (Figure 1A and Table 2). Thus the absence of substantial glucose synthesis in these segments was not due to a limitation in the provision of ATP but rather to the intrinsic organization of the glutamine metabolic pathways of rat enterocytes. It should also be mentioned here that Windmueller and Spaeth [32] have concluded from their experiments that the glutamines taken up from the lumen on the one hand and, on the other hand, from blood shared a common metabolic pool. Thus our results obtained in vitro with luminally added glutamine can be considered relevant to the in vivo situation.

The very small amounts of glucose synthesized, which were not higher in the presence than in the absence of glutamine, presumably originated from the degradation of the very small amount of glycogen present in the intestinal segments of 72 h-fasted rats (13 μmol glucosyl equivalents·entire small intestine−1).

Similarly, the absence of glutamine gluconeogenesis in our intestinal segments was not due to the re-utilization of [13C]glucose initially synthesized from [3-13C]glutamine as strongly suggested not only by the use of glucosamine, an inhibitor of hexokinase [24], but also by the incubation of [13C]glutamine with a physiological concentration (5 mM) of glucose (see Figures 1B and and2).2). Under the latter experimental condition (see the Results section), a strong line of evidence that no synthesis of [13C]glucose from [3-13C]glutamine occurred is that the resonance intensities and areas (α+β anomers) of the six glucose carbons were equal (see Figure 2 and the Results section); indeed, as recently shown in rat renal proximal tubules, gluconeogenesis from [3-13C]glutamine leads to glucose carbons labelled with carbon 13 to different degrees [21]. It should also be mentioned here that enterocytes prepared from the small intestine adapted to a high-protein diet did not produce glucose from glutamine [42].

Arteriovenous concentration difference measurements

To our knowledge, our study is the first to report arteriovenous concentration difference measurements across portal-drained viscera of 72 h-fasted rats. It should be pointed out that, unlike Mithieux et al. [5] in their in vivo experiments, we did not ligature the inferior mesenteric artery in order to avoid surgical stress. But this is unimportant with respect to glutamine metabolism because it occurs primarily in the small intestine; indeed, 87% of glutaminase is located in the small intestine [43] and therefore, as noted by Windmueller and Spaeth [32], most of the glutamine uptake in our rats occurred in the small intestine.

After 72 h of fasting, our observation of a net intestinal uptake of glucose in both Wistar and Sprague–Dawley rats (Table 3) markedly contrasts with the findings of Mithieux and co-workers [4,5] who reported no net production or uptake of glucose by the small intestine of both 48 h- and 72 h-fasted Sprague–Dawley rats. However, this is in agreement with the observation of Burrin et al. [44] who also found no evidence of gluconeogenesis in the small intestine of 36 h-fasted pigs. Since a very small, virtually negligible, glycerol uptake was observed only in Sprague–Dawley rats (Table 3), this means that the small intestine in vivo cannot synthesize substantial amounts of glucose from this substrate.

Wistar rats

Considering that two glutamine molecules are needed to synthesize one glucose molecule, and neglecting the fact that about half the circulating glutamine taken up by the intestine was oxidized to CO2, one may calculate from our data that the arterial glucose entering the intestine would have been diluted by only 1.2% (0.174:2:7.51) by the glutamine-derived glucose; the corresponding value would have been only 0.6% if one considers that half the glutamine taken up was oxidized to CO2. Unfortunately, Mithieux et al. [5] did not report the intestinal substrate uptake in their experiments. In agreement with observations of Windmueller and Spaeth [13] when the blood lactate was elevated, substantial amounts of lactate were taken up by the intestine of our 72 h-fasted Wistar rats, but this potential gluconeogenic substrate could not have been converted into glucose because of the extremely low pyruvate carboxylase activity measured in the small intestine of the adult rat [45,46]. In addition, Mithieux and co-workers [4] did not observe any conversion of [14C]lactate into [14C]glucose in the small intestine of 48 h-fasted rats in vivo. Furthermore, our preliminary results in vitro indicate that 5 mM [13C]lactate is not converted into [13C]glucose in intestinal segments prepared from 72 h-fasted Wistar rats (results not shown).

Sprague–Dawley rats

It should be pointed out that, in our 72 h-fasted Sprague–Dawley rats, glutamine was the sole potential gluconeogenic precursor taken up in substantial amounts by the intestine (Table 3). Doing the same calculations as above and neglecting the fact that some if not all of the glutamate released originated from glutamine (Table 3), it can be calculated that, if all the glutamine taken up were converted into glucose, this would have diluted the arterial glucose entering the intestinal circulation by 0.53% (0.085:2:8.00); taking into account that about half of this glutamine was converted into CO2, the corresponding value would have been only 0.27%. These values are different from that (7.9%) published by Mithieux et al. [5] in 72 h-fasted Sprague–Dawley rats. Therefore our results clearly indicate that both in 72 h-fasted Wistar rats and in 72 h-fasted Sprague–Dawley rats, glutamine gluconeogenesis cannot be detected in a reliable manner in the intestine in vivo. Since our data (Table 3) together with those of Windmueller and Spaeth [31] indicate that there were no other potential gluconeogenic precursors taken up by the intestine, this means that intestinal gluconeogenesis cannot be detected in vivo. In this context, the reasons why Windmueller and Spaeth [12,13] observed that small percentages of [U-14C]glutamine (4.7%), [U-14C]lactate (0.1–0.2%) and [U-14C]aspartate (10%) were found in [14C]glucose in the small intestine of their animals deserve careful examination. Since these authors had to chromatographically separate the various products resulting from the intestinal metabolism of their 14C-labelled substrates, their [14C]glucose fractions might have been contaminated by some other neutral 14C-labelled compounds. In this respect, we used 13C NMR spectroscopy, which avoids metabolite separations. In agreement with this view, Windmueller and Spaeth [13] found that 6.2% of 14C-labelled D-β-hydroxybutyrate was also converted into [14C]glucose; since ketone bodies cannot be gluconeogenic [47] their observation must have been due to some methodological artefact if one accepts that the small intestine does not synthesize glucose at all.

Reasons for the discrepancy between our in vivo results and those of Mithieux and co-workers

First, it should be pointed out that the arteriovenous difference approach has many limitations. Indeed, numerous experimental steps are necessary to get the final values [blood sampling, centrifugation to get the plasma, possible heterogeneity between the plasma and whole blood often neglected, deproteinization and neutralization of the sample, errors inherent to the enzymatic measurements of metabolites, imperfect chromatographic separation of radioactive metabolites, high blood concentration of glucose, radioactive measurements (radioactive disintegration is a random process), small arteriovenous differences multiplied by high blood flow sometimes measured in separate experiments]. Since each step has its own margin of error, the final values may suffer from substantial errors. Bearing this in mind, we believe that the gluconeogenic rates reported by Mithieux et al. [5] in their 72 h-fasted Sprague–Dawley rats are not compatible with the possible uptake of gluconeogenic substrates. Indeed, they calculated an intestinal gluconeogenesis of 16.9 μmol·kg body weight−1·min−1. Taking a mean body weight of 270 g and an intestinal blood flow of 5.2 ml·min−1 [5], this would mean that 0.885 (4.6:5.2) μmol of glucose was added to each millilitre of blood passing through the intestine. Since two molecules of glutamine are needed to synthesize one molecule of glucose, their results would imply that 1.77 μmol of glutamine (or of other gluconeogenic precursor) was taken up from each millilitre of arterial blood only for the synthesis of glucose and without any complete oxidation of glutamine into CO2. Our data, presented in Table 3, indicate that in fact only 0.085 μmol of glutamine was taken up from each millilitre of arterial blood in 72 h-fasted Sprague–Dawley rats (Table 3). Therefore compelling experimental evidence is provided that the gluconeogenic rates reported by Mithieux and co-workers are compatible neither with the uptake and availability of glutamine nor with the availability and uptake of other circulating gluconeogenic precursors.

Moreover, taking into account that for each molecule of glucose synthesized two glutamine carbon skeletons should pass through PEPCK, a mandatory enzyme in glutamine gluconeogenesis, it is important to note that the Vmax of PEPCK measured by Mithieux et al. [5] was severalfold lower than the rates of gluconeogenesis they calculated. This represents another compelling piece of evidence that the high gluconeogenic rates reported by the latter authors are not compatible with the biochemical characteristics of the intestine of their 72 h-fasted Sprague–Dawley rats.

Glycogen stores and glucose release in vivo

Our measurements indicate that the glycogen store in the entire small intestine of 72 h-fasted Wistar rats was 13 μmol (in glucosyl equivalents). The complete release of this glycogen as glucose via glycogenolysis would have to occur within 3 min to account for the rate of release of newly synthesized glucose reported by Mithieux et al. [5]. Thus glycogen cannot be the source of the glucose calculated to be released in vivo by these authors.

Ketone body metabolism

Another interesting observation made in the present study is that, in the 72 h-fasted rat, there were major strain differences that, to our knowledge, have never been reported so far. Indeed, the intestine of 72 h-fasted Sprague–Dawley rats did not remove circulating ketone bodies at all. By contrast, the intestine of 72 h-fasted Wistar rats removed circulating D-β-hydroxybutyrate but released approx. 60% of it as acetoacetate. Thus these two strains of rat differ from the Mendel–Osborne rats used by Windmueller and Spaeth [13] who found that, after an overnight fast, the intestine took up both D-β-hydroxybutyrate and acetoacetate. Therefore the traditional view that ketone bodies are important energy substrates of the intestine in the fasting state does not necessarily apply to all strains of rat.

In summary, our in vitro data clearly demonstrate that metabolically viable segments from the small intestine of 72 h-fasted rats did not synthesize glucose at all despite a high rate of glutamine utilization and metabolism. Our in vivo data not only in Wistar but also in Sprague–Dawley rats after 72 h of fasting indicate that the rate of intestinal gluconeogenesis reported by Mithieux et al. [5] in their 72 h-fasted Sprague–Dawley rats can be accounted for neither by the uptake of the gluconeogenic precursors available in the arterial blood, nor by the Vmax of PEPCK they measured. Moreover, our results demonstrate that glutamine gluconeogenesis would not be detectable by the currently available methods because the dilution of circulating glucose by newly synthesized glucose would be much smaller than the margins of error of the methods employed. Therefore the view that the intestine of the rat is an important site of gluconeogenesis, at least in 72 h-fasted rats, is no longer tenable.


We thank Claire Morel for secretarial assistance.


1. Nordlie R. C. Fine tuning of blood glucose concentrations. Trends Biochem. Sci. 1985;10:70–75.
2. Radziuk J., Pye S. Hepatic glucose uptake, gluconeogenesis and the regulation of glycogen synthesis. Diabetes Metab. Res. Rev. 2001;17:250–272. [PubMed]
3. Eid A., Bodin S., Ferrier B., Delage H., Boghossian M., Martin M., Baverel G., Conjard A. Intrinsic gluconeogenesis is enhanced in renal proximal tubules of Zucker diabetic fatty rats. J. Am. Soc. Nephrol. 2006;17:398–405. [PubMed]
4. Croset M., Rajas F., Zitoun C., Hurot J. M., Montano S., Mithieux G. Rat small intestine is an insulin-sensitive gluconeogenic organ. Diabetes. 2001;50:740–746. [PubMed]
5. Mithieux G., Bady I., Gautier A., Croset M., Rajas F., Zitoun C. Induction of control genes in intestinal gluconeogenesis is sequential during fasting and maximal in diabetes. Am. J. Physiol. Endocrinol. Metab. 2004;286:E370–E375. [PubMed]
6. Mithieux G., Misery P., Magnan C., Pillot B., Gautier-Stein A., Bernard C., Rajas F., Zitoun C. Portal sensing of intestinal gluconeogenesis is a mechanistic link in the diminution of food intake induced by diet protein. Cell. Metab. 2005;2:321–329. [PubMed]
7. Mithieux G., Gautier-Stein A., Rajas F., Zitoun C. Contribution of intestine and kidney to glucose fluxes in different nutritional states in rat. Comp. Biochem. Physiol. B Biochem. Mol. Biol. 2006;143:195–200. [PubMed]
8. Watford M., Lund P., Krebs H. A. Isolation and metabolic characteristics of rat and chicken enterocytes. Biochem. J. 1979;178:589–596. [PMC free article] [PubMed]
9. Baverel G., Lund P. A role for bicarbonate in the regulation of mammalian glutamine metabolism. Biochem. J. 1979;184:599–606. [PMC free article] [PubMed]
10. Rajas F., Bruni N., Montano S., Zitoun C., Mithieux G. The glucose-6 phosphatase gene is expressed in human and rat small intestine: regulation of expression in fasted and diabetic rats. Gastroenterology. 1999;117:132–139. [PubMed]
11. Rajas F., Croset M., Zitoun C., Montano S., Mithieux G. Induction of PEPCK gene expression in insulinopenia in rat small intestine. Diabetes. 2000;49:1165–1168. [PubMed]
12. Windmueller H. G., Spaeth A. E. Metabolism of absorbed aspartate, asparagine, and arginine by rat small intestine in vivo. Arch. Biochem. Biophys. 1976;175:670–676. [PubMed]
13. Windmueller H. G., Spaeth A. E. Identification of ketone bodies and glutamine as the major respiratory fuels in vivo for postabsorptive rat small intestine. J. Biol. Chem. 1978;253:69–76. [PubMed]
14. Hahn P., Wei-Ning H. Gluconeogenesis from lactate in the small intestinal mucosa of suckling rats. Pediatr. Res. 1986;20:1321–1323. [PubMed]
15. Mithieux G., Rajas F., Gautier-Stein A. A novel role for glucose 6-phosphatase in the small intestine in the control of glucose homeostasis. J. Biol. Chem. 2004;279:44231–44234. [PubMed]
16. Watford M. Is the small intestine a gluconeogenic organ? Nutr. Rev. 2005;63:356–360. [PubMed]
17. Chauvin M. F., Mégnin-Chanet F., Martin G., Lhoste J. M., Baverel G. The rabbit kidney tubule utilizes glucose for glutamine synthesis. A 13C NMR study. J. Biol. Chem. 1994;269:26025–26033. [PubMed]
18. Chauvin M. F., Mégnin-Chanet F., Martin G., Mispelter J., Baverel G. The rabbit kidney tubule simultaneously degrades and synthesizes glutamate. A 13C NMR study. J. Biol. Chem. 1997;272:4705–4716. [PubMed]
19. Dugelay S., Chauvin M. F., Megnin-Chanet F., Martin G., Lareal M. C., Lhoste J. M., Baverel G. Acetate stimulates flux through the tricarboxylic acid cycle in rabbit renal proximal tubules synthesizing glutamine from alanine: a 13C NMR study. Biochem. J. 1999;342:555–566. [PMC free article] [PubMed]
20. Conjard A., Dugelay S., Chauvin M. F., Durozard D., Baverel G., Martin G. The anaplerotic substrate alanine stimulates acetate incorporation into glutamate and glutamine in rabbit kidney tubules. A 13C-NMR study. J. Biol. Chem. 2002;277:29444–29454. [PubMed]
21. Vercoutère B., Durozard D., Baverel G., Martin G. Complexity of glutamine metabolism in kidney tubules from fed and fasted rats. Biochem. J. 2004;378:485–495. [PMC free article] [PubMed]
22. Vittorelli A., Gauthier C., Michoudet C., Martin G., Baverel G. Characteristics of glutamine metabolism in human precision-cut kidney slices: a 13C-NMR study. Biochem. J. 2005;387:825–834. [PMC free article] [PubMed]
23. Baverel G., Martin G., Ferrier B., Vercoutère B., Saadé F., Conjard A., Delage H., Boghossian M., Vanbelle C., Chauvin M. F., et al. In vitro, la glutamine est un précurseur de glucose dans les cellules hépatiques et rénales mais pas dans les cellules intestinales de rats nourris, à jeun et diabétiques. Diabetes Metab. 2005;31:1577–1578.
23a. Krebs H. A., Henseleit K. Untersuchungen uber die Harnstoffbildung im Tierkörper. Hoppe-Seyler's Z. Physiol. Chem. 1932;210:33–66.
24. Guglielminetti L., Perata P., Morita A., Loreti E., Yamaguchi J., Alpi A. Characterization of isoforms of hexose kinases in rice embryo. Phytochemistry. 2000;53:195–200. [PubMed]
25. Lamprecht W., Trautschold I. Adenosine-5′-triphosphate. Determination with hexokinase and glucose-6-phosphate dehydrogenase. In: Bergmeyer H., editor. Methods of Enzymatic Analysis. New York: Academic Press; 1974. pp. 2101–2110.
26. Conjard A., Brun V., Martin M., Baverel G., Ferrier B. Effect of starvation on glutamine ammoniagenesis and gluconeogenesis in isolated mouse kidney tubules. Biochem. J. 2002;368:301–308. [PMC free article] [PubMed]
27. Passonneau J. V., Lowry O. H. Enzymatic Analysis: A Practical Guide. Totowa, NJ: The Humana Press; 1993.
28. Howarth O. W., Lilley D. M. J. Carbon-13-NMR of peptides and proteins. Prog. NMR Spectrosc. 1978;12:1–40.
29. Canioni P., Alger J. R., Shulman R. G. Natural abundance carbon-13 nuclear magnetic resonance spectroscopy of liver and adipose tissue of the living rat. Biochemistry. 1983;22:4974–4980. [PubMed]
30. James L. A., Lunn P. G., Elia M. Glutamine metabolism in the gastrointestinal tract of the rat assessed by the relative activities of glutaminase (EC and glutamine synthetase (EC Br. J. Nutr. 1998;79:365–372. [PubMed]
31. Windmueller H. G., Spaeth A. E. Uptake and metabolism of plasma glutamine by the small intestine. J. Biol. Chem. 1974;249:5070–5079. [PubMed]
32. Windmueller H. G., Spaeth A. E. Intestinal metabolism of glutamine and glutamate from the lumen as compared to glutamine from blood. Arch. Biochem. Biophys. 1975;171:662–672. [PubMed]
33. Dickens F., Weil-Malherbe H. Metabolism of normal and tumour tissue. Biochem. J. 1941;35:7–15. [PMC free article] [PubMed]
34. Anderson J. W. Glucose metabolism in jejunal mucosa of fed, fasted, and streptozotocin-diabetic rats. Am. J. Physiol. 1974;226:226–229. [PubMed]
35. Ardawi M. S. Glutamine and ketone-body metabolism in the gut of streptozotocin-diabetic rats. Biochem. J. 1988;249:565–572. [PMC free article] [PubMed]
36. Brosnan J. T., Man K. C., Hall D. E., Colbourne S. A., Brosnan M. E. Interorgan metabolism of amino acids in streptozotocin-diabetic ketoacidotic rat. Am. J. Physiol. 1983;244:E151–E158. [PubMed]
37. Watford M., Smith E. M., Erbelding E. J. The regulation of phosphate-activated glutaminase activity and glutamine metabolism in the streptozotocin-diabetic rat. Biochem. J. 1984;224:207–214. [PMC free article] [PubMed]
38. Welbourne T. C., Phromphetcharat V., Givens G., Joshi S. Regulation of interorganal glutamine flow in metabolic acidosis. Am. J. Physiol. 1986;250:E457–E463. [PubMed]
39. Kong S. E., Hall J. C., Cooper D., McCauley R. D. Starvation alters the activity and mRNA level of glutaminase and glutamine synthetase in the rat intestine. J. Nutr. Biochem. 2000;11:393–400. [PubMed]
40. McManus J. P., Isselbacher K. J. Effect of fasting versus feeding on the rat small intestine. Morphological, biochemical, and functional differences. Gastroenterology. 1970;59:214–221. [PubMed]
41. Zager R. A. Adenine nucleotide changes in kidney, liver, and small intestine during different forms of ischemic injury. Circ. Res. 1991;68:185–196. [PubMed]
42. Azzout-Marniche D., Gaudichon C., Blouet C., Mathe V., Bos C., Huneau J. F., Tomé D. Glucoregulation in rat adapted to a high protein diet. FASEB J. 2005;19:A74.
43. Pinkus L. M., Windmueller H. G. Phosphate-dependent glutaminase of small intestine: localization and role in intestinal glutamine metabolism. Arch. Biochem. Biophys. 1977;182:506–517. [PubMed]
44. Burrin D., Lambert B., Stoll B., Guan X. Gastrointestinal gluconeogenesis is absent in the 36-hour fasted piglet. FASEB J. 2005;19:A1675.
45. Anderson J. W. Pyruvate carboxylase and phosphoenolpyruvate carboxykinase in rat intestinal mucosa. Biochim. Biophys. Acta. 1970;208:165–167. [PubMed]
46. Hahn P., Taller M., Chan H. Pyruvate carboxylase, phosphate-dependent glutaminase and glutamate dehydrogenase in the developing rat small intestinal mucosa. Biol. Neonate. 1988;53:362–366. [PubMed]
47. Krebs H. A., Hems R., Weidemann M. J., Speake R. N. The fate of isotopic carbon in kidney cortex synthesizing glucose from lactate. Biochem. J. 1966;101:242–249. [PMC free article] [PubMed]

Articles from Biochemical Journal are provided here courtesy of The Biochemical Society
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • Compound
    PubChem Compound links
  • MedGen
    Related information in MedGen
  • PubMed
    PubMed citations for these articles
  • Substance
    PubChem Substance links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...