recycling of glucose by rat hepatocytes

Em. J. Biochem. 60, 91- 101 (1975)

Recycling of Glucose by Rat Hepatocytes Joseph KATZ, P. A. WALS, Sybil GOLDEN, and Robert ROGNSTAD

Cedars-Sinai Medical Center, Los Angeles, California

(Received May 27/August 19, 1975)

1. The metabolism of glucose labeled uniformly with I4C, and in positions 2, 3 and 5 with tritium by hepatocytes from fed and fasted rats were studied. Cells were incubated with glucose as sole substrate, or with glucose and a variety of glucose precursors, and uptake or production of glucose, and the utilization of the isotopes was determined.

2. There was no uptake of glucose at concentration of up to 15 mM, and net glucose synthesis in the presence of precursors. I4C was however recovered in COz, lactate and amino acids, and tritium in water. Considerable incorporation into glycogen from '"C and 3H-labeled glucose occurred at high (above 20 mM) glucose concentrations.

3. The yield in water always exceeded that in I4C-labeled products. The yield in 3HOH from [2-3H]glucose exceeded that from [5-3H]glucose, and the latter was greater than from [3-3H]glucose.

4. Utilization of labeled glucose does not follow Michaelis-Menten kinetics. The fractional rate of uptake of 14C and tritium-labeled glucose increases with glucose concentration with a maximum at about 15 mM and then declines.

5. The effect of numerous gluconeogenic substrates on the isotope utilization and the 3H/14C ratio in glycogen was studied. The uptake of I4C was always depressed. Addition of lactate and dihydroxyacetone has little effect on the detritiation of [2-3H]glucose, but it is depressed by other substrates. The detritiation of [3-3H]- and [5-3H]glucose is depressed in gluconeogenesis, that from [3-3H]glucose usually more than from [5-3H]glucose. In the presence of lactate detritiation of [3-3H]glucose is about half that from [5-3H]glucose.

6. Equations to calculate the phosphorylation of glucose and fructose 6-phosphate in the presence of futile cycling between glucose and glucose 6-phosphate and fructose 6-phosphate and fructose 1,6-bisphosphate were derived.

7. The estimate of glucose phosphorylation requires determination of the specific activity of glucose 6-phosphate from [2-3H]glucose. It appears that futile cycling between glucose and glucose 6-phosphate is extensive in cells with a high glycogen content, but is low in cells from starved rats and nearly absent in those from diabetic animals.

8. The estimation of the phosphorylation of fructose 6-phosphate in the presence of cycling requires knowledge of the specific activities of fructose 6-phosphate and fructose 1,6-bisphosphate from [3-3H]glucose. At present there are no adequate data to calculate phosphorylation and recycling of fructose 6-phosphate, but under some conditions the rate may be quite high.

Liver contains the full complement of enzymes for glucose synthesis and for glycolysis. Two irreversible steps in glucose metabolism are between glucose and glucose-6-P, and between fructose-6-P and fmctose- 1,6-P2. If the kinases and phosphatases catalyzing these conversions were simultaneously active, there would be two futile cycles, namely glucose + glucose- 6-P + glucose (the glucose cycle), and fructose-6-P + fructose-l,6-P2 + fructose-6-P (the fructose 6-P cycle). Newsholme and coworkers [l - 31 have sug-

Abbreviations. Glucose-6-P, glucose 6-phosphate; fructose-6-P, fructose 6-phosphate; fructose 1,6-P,, fructose 1,6-bisphosphate; glyceraldehyde-3-P, glyceraldehyde 3-phosphate.

gested that futile cycles have a role in metabolic regulation. The fructose-6-P cycle serves in thermogenesis in flight muscle of bumble bees [4] and possibly mammalian muscle [5]; a similar function has been suggested for the fructose-6-P cycle in liver [6]. Experimental support for the recycling of glucose in liver was obtained recently. Evidence for the glucose cycle was provided in rat hepatocytes with [Z3H, U-14C]glucose [7,8], and for the fructose-6-P cycle with [5-3H, U-'4C]glucose [7,9]. Hue and Hers [lo] confirmed the occurrence of the glucose cycle in mouse liver, but questioned the existence of significant futile cycling between fructose-6-P and fructose-l,6-Pz.

92 Glucose Recycling by Hepatocytes

The physiological role of futile cycles and their function in metabolic regulation is of great interest. Also evaluation of recycling is essential to estimate the true rates of phosphorylation of glucose and fructose-6-P and of the glucose 6-phosphatase and fructose 1,6-phosphatase in intact cells. Methods to establish and quantitate recycling are based on the use of tritium-labeled glucoses and interpretation of this type of isotopic data offers difficulties. We thus set out to study further the metabolism of tritium-labeled glucoses by hepatocytes from rats in different dietary states, and to examine the validity of the methods proposed for the quantitation of recycling. Our conclusion is that there is extensive recycling between glucose and glucose-6-P in hepatocytes of fed rats. There is also recycling at the fructose-6-P level but its extent is at present difficult to estimate.

MATERIALS AND METHODS

Animals

Rats of the Wistar strain, 170-250 g in weight, were either fasted for 20 h or fed a commercial pellet diet (high in starch) or a 60% sucrose, low fat diet sold by Nutritional Biochemical Co. (Cleveland, Ohio). Access to food was either ad libitum or was restricted, by means of an automatic device to 8 - 10 a.m. (meal fed rats). Diabetes was induced by a sub- cutaneous injections of streptozotocin, 100 mg/kg.

Hepatocytes

The preparation was according to Clark et al. [ l l ] except for modifications described below. Krebs- Henseleit buffer, without calcium, equilibrated with 95% 0,-5% CO,, 5 mM in glucose and 5 mM in pyruvate was used instead of the Hanks buffer. The liver was perfused by gravity in situ, (without recircula- tion) with 150 ml of the above buffer at about 35- 38 "C, at a rate of 40 - 50 ml/min and the liver detached during this procedure. The liver was transferred to a perfusion assembly and 120 ml buffer containing 0.4 mg/ml of collagenase (Worthington) recirculated at a rate of 80- 100 ml per min. Gas (95 % O2 - 5 % C02) was bubbled through the buffer prior to its entry into the liver. After about 20 min the liver was gently dispersed and the cells washed 4 to 5 times with buffer containing calcium, but without glucose and pyruvate. The yield of packed cells ranged from 4- 5 ml from a liver of a 200 g fasted rat to 8 - 10 ml from a starved rat refed a high sucrose diet.

Labeled Sugars

These were purchased from Amersham-Searle (Chicago, Illinois) or New England Nuclear (Boston,

Massachusetts). The I4C sugars were found to be of good quality, but the purity of tritium-labeled sugars was found to vary from batch to batch from the same manufacturer. Radioactive contamination in tritium- labeled compounds is difficult to detect by radio- autography, and purification by paper chromatog- raphy was not always successful. Therefore the tritium- labeled glucose was mixed with [U-14C]glucose, a few pmoles of carrier added and the glucose converted enzymatically to glucose-6-P. The latter was adsorbed on a small column of Dowex-1 acetate resin. All the sugar in the mixture was quantitatively converted to glucose-6-P, and 100% of the I4C adsorbed on the column but varying amounts, up to 30% of tritium- labeled material, was not phosphorylated and not adsorbed on the column. The glucose-6-P was eluted with 4N formic acid, the acid evaporated off, and the glucose regenerated with acid phosphatase. The solution was deproteinized, deionized through a mixed-bed ion-exchange resin and taken to dryness. A number of the batches of the sugars were degraded [12] to check on the specificity of labeling. In several prepara- tions, between 5 to 20 % of the tritium was not in the specified position. From our experience we believe that the purity of commercial tritiated sugars and the specificity of labeling is variable and that caution is required in the interpretation, especially when dealing with low isotopic yields in the products.

Incubations

The cells were incubated in the Krebs-Henseleit buffer, which in some experiments contained 1.5 % gelatine [13]. We observed that gelatine decreases clumping and adhesion of cells and increases somewhat the utilization of substrates. The incubations were terminated with perchloric acid (6 % final concentration) which does not precipitate gelatine.

Isolation of Labeled Compounds

In the main this was as previously described [ l l ] except for the collection of labeled water and isolation of glycogen. Glycogen is completely extracted with perchloric acid. The glycogen was precipitated from an aliquot of the perchloric acid extracts with 2 volu- mes of 95% ethanol. After standing overnight in the cold the glycogen was centrifuged down, redissolved in 0.5 ml 1 N NaOH, and reprecipitated from a volume of 10 ml of 65% ethanol. The precipitate was washed thoroughly with aqueous ethanol, redissolved in 0.1 N sodium acetate buffer pH 4.5, containing 1 mg/ml amyloglucosidase (Sigma). After 2 h incubation at 55 "C, aliquots of the solution were assayed for radioactivity and for glucose.

An aliquot of the perchloric acid extract was neutralized with KOH, the salt removed and the

J. Katz, P. A. Wals, S. Golden, and R. Rognstad 93

solution passed through three small (0.8 x S cm) tandem columns of Dowex-1 (H’, 50- 100 mesh), Dowex SO acetate (100-200 mesh) and Dowex 50 borate (100-200 mesh). The eluate contains the labeled water. The glucose is absorbed on the borate column and was eluted with 0.5 N acetic acid. The other fractions were obtained as previously described [ll]. For each isotope combination a zero-time control was processed through the entire fractionation, and the I4C and tritium blanks recovered in each fraction subtracted from the results. Since yields were low under some conditions the corrections for blanks are significant.

Expressing the Results

Investigators have used a great variety of methods to express their results with hepatocytes, per number of cells, per mg protein, on a wet or dry basis. The methods are not easily interconvertible, nor is comparison with intact liver readily possible. The number of cells per volume decreases with glycogen content. We found that 1 ml of packed volume of hepatocytes obtained from fasted rats contained 145 x lo6 cells, 110 x lo6 cells from ad libitum meal-fed rats, and from 60- 90 x lo6 cells/ml in meal-fed or fasted-refed rats. The “wet weight” of cells is not equivalent to that of intact liver. We express our results on the basis of the perchloric-acid-insoluble, glycogen-free, defat- ted dry weight. This fraction is probably 95 % protein. To obtain this fraction cells were treated with perchloric acid in tared tubes and the residue extracted with chloroform/methanol and the residue dried at 60 “C to constant weight. in some experiments every vessel was processed in this way and the dry weight (20 to 30 mg) was found to be highly uniform. Our results are expressed per 100 mg of this dry weight designated here as protein. The weight of the liver, and the water and protein content of packed hepatocytes as well as of intact liver changes rapidly with the diet. We observed 29% protein (as percentage wet weight) in livers of fasted rats, but only 15 % in meal- fed rats with enlarged livers with a 10- 12 % glycogen content. Upon refeeding a high sucrose diet to the fasted rat the liver was nearly doubled in weight in a day or two but the change in total protein was small. In the experiments of this paper the number of cells ranged between 100 to 140 x lo6 per ml packed cells. As a rough approximation our results (expressed per 100 mg of protein) can be converted to “wet weight” by multiplying by a factor of 2.5-3 for cells low in glycogen, and by a factor of 2 for livers high in glycogen.

The Fate of Tritium from Glucose

Since this study of the glucose and fructose-6-P cycles is altogether based on the loss of labeled hydro-

gens from glucose, a short exposition of the fate of tritium in glycolysis is in order. Hydrogen from position 2 of glucose is transferred in the isomerization of glucose-6-P to position 1 of fructose-6-P. However in a single transfer a variable fraction, up to one half the hydrogen exchanges with protons of the medium [14]. i f the rate of hexose-6-P isomerase is very much more rapid than the net flux, the tritium from position 2 of glucose would be completely lost from glucose-6-P, and the yield of labeled water would provide a true measure of glucose phosphorylation. However, it is not likely that the rate of phosphohexose isomerase is rapid enough to cause complete detritiation at the glucose-6-P level [15]. If the loss at the glucose-6-P stage is incomplete and there is recycling, [2-3H]glucose-6-P would be reconverted to glucose, and the yield in water would not provide a valid estimate of glucose phosphorylation and recycling. The yield of 3HOH would provide then only the minimal value for glucose phosphorylation and the true rate will be underestimated.

In the aldolase cleavage of fructose-1,6-P2 the hydrogen from position 3 of glucose appears on C-1 of dihydroxyacetone phosphate and that from position 5 in position 2 of glyceraldehyde phosphate. In the triose phosphate isomerase reaction the tritium from these positions is nearly completely exchanged with protons [14] and would appear in water. if the triose phosphate isomerase were much more rapid than the net flux and the aldolase reaction readily reversible and also very rapid, fructose-1 ,6-Pz from either [3-3H]- or [S-3H]glucose would be virtually tritium-free. Under these conditions, if there is no transaldolase catalyzed exchange between fructose- 6-P and glyceraldehyde 3-P (see below) the formation of labeled water from the two tracers would be equal and could be used to calculate the rate of fructose-6-P phosphorylation. If however aldolase is not extremely rapid (even if detritiation in the triose phosphate iso- nierase reaction were complete), (see Appendix) fructose-1 ,6-P2 will contain tritium and would be recycled to hexose-6-P and glucose. The yield in water would thus be less than if there were no tritium retention in fructose- 1,6- Pz .

RESULTS

Typical results with hepatocytes from diabetic and fasted rats and those fed a rat chow and a high sucrose diet are shown in Table 1. Hepatocytes from fasted or fed rats do not take up glucose from the medium at concentrations below 15 -20 mM. Limited glucose utilization from the medium was seen, but not consistently, at about 15 mM by hepatocytes from fed rats. There was however considerable utilization of glycogen and net carbohydrate uptake in these cells. In the presence of low concentrations of lactate


Table 1. Carbohydrate balance and apparent utilization 0fI4C- and tritium-labeled glucose by hepatocytes from rats in dgferent conditions Cells, 10 - 30 mg protein, incubated for 1 h in 2 ml bicarbonate buffer. Label in glucose was (2-3H, U-14C) and (S3H, U-14C). Results expressed per 100 mg cell protein. Standard deviation given when it exceeded 2 15 "/,. Apparent utilization is obtained by multiplying the fraction of iso,tope recovered in total products (CO,, lactate, amino acids, lipids, glycogen from [U-14C]glucose, and water from [2-3H]- and [5-3H]glucose), by the mean glucose concentration, adjusted to 100 mg cell protein. n.d., not determined

Condition or Initial Substrate A A A A Apparent utilization diet glycogen Glycogen Glucose Carbohydrate Lactate

glucose lactate U-14C z 3 H 5-3H

Meal-fed

High sucrose ad libitum

Regular

Fasted

ad libitum

Fed streptozotacin diabetic

pno1/100 mg mM

170 1s 1s

150 15 1s

90 1s 1s

6 1s 1s

4 1s 1s

pmol x 100 mg x h-'

- - 5 s -12 & 3 -61 10 - 49 +11 & 4 -38

- - 36 - 3 2 1 -39 10 - 33 +15 - 1s

- - 24 + 2 , 2 -22 10 - 22 + 8 , 2 -15

- n.d. + 3 + 2 0 10 n.d. + 17 + 11

- - 3 f 2 2 1 0 10 - 3 + 23 + 20

- 26 3 1 21 - 127 14 35 16

- 21 38 25 -111 15 35 18

- 1.4 15 13 - 51 2.8 14 10

- 4.6 11 7.9 - 43 1.6 9.0 5.2

- 0.8 2.1 0.7 - 65 0.5 1.4 0.6

(2 mM) or other precursors, there was net synthesis of glucose. 14C from glucose was however always recovered in CO,, lactate or other products (see below), and water was formed from tritiated glucose in the absence of net glucose uptake or when glucose was synthesized. Thus the isotope yields from glucose do not represent catabolism of glucose, and such yields are designated as apparent uptake or utilization. This is calculated by multiplying the fraction of isotope recovered in total products by the mean glucose concentration during the incubation, and such results are reported for three types of labels in the last columns of Table 1. These apparent rates of [2-3H]- and [5-3H]glucose uptake represent respectively the minimal rates of phosphorylation of glucose and glucose-6-P.

It is shown in Table 1 that in cells high in glycogen there is a considerable net utilization of carbohydrate, even when glucose was formed. The fate of the carbohydrate is not established. The apparent utilization of 14C-labeled glucose is highest with rats meal-fed or kept on high sucrose diets. It was depressed in the presence of lactate, especially in rats fed ad libitum and fasted rats. The apparent uptake of labeled glucose was low in cells from fasted and even lower from streptozotacin-diabetic rats. Hepatocyte from such animals had a very high rate of glucose synthesis from lactate (not shown). Glucokinase is greatly depressed under these conditions, and this is likely to account for the low rate of apparent uptake and of recycling. The low yields of labeled water make studies with such cells difficult, and in most of our work, hepatocytes from fed rats were used.

The detritiation from [2-3H]glucose in cells of fed rats exceeded the utilization of [14C]glucose by 1.5 to 3 times. The addition of lactate depressed apparent utilization of ['"C]glucose but had little effect on that of [2-3H]glucose so that detritiation was from 3 to 5 times the utilization of 14C. The detritiation of [5-3H]glucose was lower than that of [2-3H]glucose and was also depressed by the addition of lactate.

The Kinetics of Glucose Utilization

The metabolism of glucose labeled with I4C and tritium in positions 2 and 5 by hepatocytes of rats kept on a sucrose diet was studied in more detail for a period up to 4 h. There was no significant change in glucose concentration (about 12 mM) although there was glycogen breakdown. There was very rapid initial formation of lactate, which attained a concentration of about 1 - 1.5 mM. The composition of the 14C-labeled products changed markedly with time as shown in Fig. 1 A. In the first half hour lactate was the major labeled product. Maximal yield was attained very rapidly, but did not change thereafter, whereas 14C0, production was nearly linear for a 4-11 period. There was initial rapid 14C incorporation into amino acids, but the rate declined after 1 h. (Alanine is the predominant labeled amino acid.) There was no ammonia added, so that this labeling is likely to be mainly by exchange through transamina- tion with endogenous amino acids.

In Fig. 1B the apparent utilization of [U-'"CI-, [2-3H]- and [5-3H]glucoses are compared. The ap-


160 0 0 120

v

r 0 .-

Amino acids ;;w Acids Residue

10 0 c L ip id

' f o

.- 0

Glycogen

0 1 2 3 4 Time [ h )

- 40 ," -

I m L 10

l B [2-3H]Glucose

0 [5-3H]Glucose f

0

2 0 0 1 2 3 4

Time ( h )

parent utilization of U-'"C declined somewhat faster than that of tritium. In Fig. 1C the 3H/14C ratio in glycogen and in residual glucose is shown. The 3H/14C ratio in glycogen from [5-3H, U-14C]glucose was nearly constant for 3 h, about 0.8, but that from [Z3H, U-14C]glucose declined in 3 h from 0.5 to 0.4, probably reflecting the decline in the -'H/14C ratio in substrate glucose.

Effect of Glucose Concentrntion

The apparent change in glucose and glycogen utilization of [U-'"CI-, [2-3H]- and [5-3H]glucose were studied at glucose concentrations from 3 to 60 mM, with cells of rats fed a high sucrose diet. Fig. 2A shows that at concentrations of up to 20 mM there was an increase in glucose, but at higher concentrations it was taken up. Since the change in concentration is only a few percent, measurement of uptake is subject to a large error. The best estimate is 20- 30 pmol x 100 mg-' x h-' at concentrations of 40 to 60 mM. There was however carbohydrate utilization at the expense of glycogen. The initial glycogen level was about 8 % of wet weight. The breakdown of glycogen was decreased with increasing glucose concentrations, and at 60 mM it was virtually unchanged (Fig. 1A).

In Fig. 2B and 2C the apparent utilization of the 3 labeled sugars is compared. Results are presented

0.3 0 1 2 3 4

Time (h )

Fig. 1. Kinetics of glucose metabolism by rat hepatocytes. Cells from a fed rat, about 0.1 ml packed volume, 18 mg cell protein, were incubated in 2 ml Krebs bicarbonate buffer, 15 mM in glucose at 37 "C, in an atmosphere of 95 % 0,-5% CO,. The glucose was labeled uniformly with 14C, and with tritium in either position 2 or 5. (A) Apparent incorporation of 14C into products, yields expressed as patoms of carbon per 100 mg protein. Lower figure has an enlarged scale from the upper one. The acid fraction consists mainly of lactate. (B) Apparent recovery of isotope in products from [U-14C] ; [2-3H]- and [5-3H]glucose. Water and glycogen constituted 95 % or more of the tritium labeled products. (C) Solid lines, 3H/14C ratios in medium glucose at end of period; (m) (5-3H)/(U-'4C), (0) (2-3H)/(U-'4C). Broken lines, 3H/14C ratios in glycogen; (A) (5-3H)/(U-14C), (+) (2-3H)/(U-'4C)

in two ways, as fractional rates (as percentage of the added isotope recovered in products, Fig. 2B), and as the apparent glucose incorporation (pmol x 100 mg-' x h-l, Fig. 2C). The interesting aspect of the data is that the fractional rates of uptake of the three tracers increased to a maximum at about 15 mM and then declined. In Michaelis-Menten kinetics fractional rates decline with concentration and the curves show no maxima.

In Fig. 2B and 2C the incorporation of [U-'"CI- glucose into glycogen is also shown. The curve is sigmoidal. At concentrations below 20 mM the 14C yield in glycogen is negligible, but at higher concentrations it accounts for a major fraction of the I4C yield. High concentrations of glucose activate glycogen synthase [16] but the incorporation of I4C occurs in the absence of any net glycogen synthesis. It is not an exchange between free glucose and the terminal glucose units in the glycogen chains, since the 3H/'4C ratio in glycogen is well below that in medium glucose (see Fig. 2D). However a phosphorylase catalyzed exchange between glucose-1-P and glycogen is possible. Such an exchange does not dissipate ATP and is not a futile cycle.

In Fig. 2D the ratios of uptake of the three labeled glucoses as a function of glucose concentration are shown. At low concentrations detritiation is much higher than U-14C uptake but it declines to constant values at about 15 mM. The 3H/14C ratios in glycogen


0 10 20 30 40 5 0 60 Mean [glucose] ( m M )

(5-3H)

( ~ - ' ~ ~ ) t o t a \

( u - ' ~ c ) glycolysis

[ U-14C] Glycogen

0 10 20 30 40 50 60 Mean [glucose] (mM)

\ (5-3H)/(U- '4C)

( 2 - 3 H ) / ( 5 - 3 H )

, ,-l4c in glycogen

[UJ4C]Glucose

, , I

0 10 20 30 40 50 60 Mean [glucose] (mM)

1

$ l o -

y 5 -

0 0 10 20 30 40 50 60

Mean \glucose] (mM)

Fig.2. Effect of glucose concentration on the utilization of labeled glucose. Cells from a rat fed a high sucrose diet, 15 mg cell protein (about 80 mg wet weight) were incubated for 1 h in 1.5 ml Krebs-Henseleit bicarbonate buffer with glucose labeled with 14C and with tritium in positions 2 or 5. (A) Tissue glycogen and glucose uptake or production, and net carbohydrate utilization as determined by analysis. Results expressed as pmol per flask. (0) Tissue glycogen; (0) production of glucose; (+)net carbohydrate uptake; (A) [U-14C]glucose utilisation. (B) The recovery of isotope (as % added 14C or 3H) in total products and of 14C in glycogen. (C) Apparent utilization of labeled glucose. Glycolysis represents the 14C yield in all products except glycogen. Apparent yield is the product of the fraction of added isotope recovered in the products and mean glucose concentration. (D) Upper curves, the ratio of isotope recovered in total products. Lower curves, 'H/14C ratios in glycogen from [Z3H, U-14C]- and [S3H, U-'4C]glucose. The initial ratio in glucose is taken as 1.0

from [5-3H, U-'4C]glucose and [Z3H, U-'4C]glucose are also shown. The 5-3H/U-14C ratio remained nearly constant at 0.8, but the 2-3H/U-14C ratio in glycogen increased from about 0.3 at 10 mM glucose to about 0.5 at 60 mM.

Effect of Gluconeogenesis on the Detvitiation of [2-3HJGlucose

We have shown in Table 1, that the gluconeogenesis from added lactate depressed the apparent utilization of 14C glucose but had little effect on the yield of 3HOH from [2-3H]glucose. The effects of gluconeogenesis from a number of other substrates on the detritiation of [2-3H]glucose and the 3H/14C ratio in glycogen was studied. The results shown in Table 2, indicate that the effect depends on the nature of the substrate. All compounds tested except dihydroxyacetone greatly depressed apparent utilization of 14C glucose. Dihydroxyacetone was unique in that the depression of 14C uptake was much less than by other substrates. Lactate and dihydroxyacetone had only small, barely significant effects of the detri-

tiation of [2-3H]glucose, but pyruvate, fructose and sorbitol were inhibitory. The depression in the presence of xylitol, glycerol and glyceraldehyde was most pronounced. Of special interest is the effect of the addition on the 3H/'4C ratio in glycogen from p 3 H , U-14C]glucose. The ratio was not much affected by xylitol, depressed by glycerol, lactate and pyruvate, and elevated in the presence of fructose, sorbitol and dihydroxyacetone. With fructose in cells from fasted rats ratios as high as 0.8-0.9 were observed. The increased retention of hydrogen in glucose-6-P is probably due to an inhibition of phosphohexose isomerase by fructose-l-P [17]. Fructose and sorbitol were found to depress the incorporation of 3HOH into position 2 of glucose synthesized by hepatocytes [18].

When the .retention of tritium in position 2 of glucose-6-P is high, most of the label will return to glucose and only a part liberated as water. Thus in the presence of fructose and sorbitol, recycling between glucose and glucose-6-P could be equal or even higher than that with lactate in spite of the depressed yield in water. It appears that the highest recycling occurs in the presence of rapid gluconeogenesis.


Table 2. Effect of additions on the utilization 0 f [ 2 - ~ H , U-’4C]g/ucose and the 3H/’4C ratio in glycogen Each group average o f 2 experiments, and results expressed as pmol x (100 mg cell protein)-’ x h-’. 15-25 mg of cell protein, incubated for 40 niin in 4 ml bicarbonate buffer, 10 mM in glucose and other substrates. The incorporation in the presence of glucose as sole substrate is set to 100 %

Diet Added substrate A A Apparent incorporation of [Z3H, U-14C]glu~ose in 3H/’4C Glucose Glycogen ratio in

co2 water glycogen

pmol x 100 mg-l pmol x 100 mg-’ p m o ~ x 100 mg-’ % x h-’ x h-’ x h-’

Fasted None Lactate Pyruvate Fructose Sorbitol Dihydroxyacetone Xylitol Glycerol Glyceraldehyde

Lactate Pyruvate Fructose Sorbitol Dihydroxyacetone Xylitol Glycerol Glyceraldehyde

Fed None

+ 3 + 23 + 24 + 81 + 39 + 33 + 47 + 30 + 30

+ I + 24 + 25 + 64 + 51 + 43 + 43 + 30 + 65

1.6 100 12 16 16 48 39 73 26 23 26

- 42 2.9 - 32 - 30 - 10 - 9 - 14 - 30 - 27 - 41

100 18 27 26 48 27 79 22 20 21

100 0.36 89 0.43 67 0.58 48 0.92 78 0.84

109 0.48 59 0.67 50 0.49 23 0.80

100 0.38 94 0.29 78 0.25 62 0.55 85 0.59

105 0.40 51 0.47 50 0.37 35 0.44

The Role of Transaldolase in the Detritiation of [5-3HJGlucose

Clark et al. [9] assumed that labeled glyceraldehyde 3-P is formed solely by aldolase cleavage of fructose- 1,6-P2. Liver however has an active transaldolase. This enzyme catalyzes an exchange between the 3 “bottom” carbons of fructose-6-P and free glyceraldehyde 3-P. Tritium from [5-3H]fructose would thus appear in position 2 of glyceraldehyde 3-P and would be exchanged with protons by the action of triose phosphate isomerase. This would constitute an exchange, without energy dissipation, rather than futile cycling which causes ATP hydrolysis. The tritium from position 5 may appear in triose phosphate by either aldolase cleavage or transaldolase exchange between fructose-6-P and glyceraldehyde-3-P, but tritium from position 3 of glucose appears in triose phosphate only after cleavage of fructose-1 ,6-P2 by aldolase. Hue and Hers [lo] compared the detritiation of these two labels in mouse liver extracts and their results showed that most of the detritiation of [5-3H]- glucose was catalyzed via transaldolase rather than via aldolase cleavage.

Results with perfused liver [I91 from fed or fasted rats shows that detritiation of [5-3H]glucose is considerably greater than that of [3-3H]glucose. However the 3HOH yield from [3-3H]glucose [I91 exceeded that

from [‘“CI- or [6-3H]glucose suggesting the operation of a futile cycle between fructose-6-P and fructose- 1,6-P2. Comparison of [3-3H]glucose and [5-3H]- glucose is complicated in the presence of the pentose cycle since [3-3H]glucose forms labeled NADPH, and the fate of the latter in liver is not well established. In Table 3 the detritiation of [3-3H]- and [5-3H]glucose by rat hepatocytes from fasted rats in the presence of a series of gluconeogenic precursors is compared. With glucose as sole substrate the yield of water from [3-3H]gl~~cose ranged from 65 to 80 % that from [5-3H]- glucose. The results in the presence of added substrates are variable, but some obvious conclusions stand out. All substrates with the exception of dihydroxyacetone depressed the utilization of [14C]glucose, as indicated by the ’“C02 yield from [U-14C]glucose, and also the yields in water. Dihydroxyacetone was unique in that it only slightly depressed the 14C02 yields and actually somewhat stimulated detritiation from both [3-3H]- and [5-3H]glucose. The different response to dihydroxyacetone and glyceraldehyde is striking. Both serve nearly equally well as precursors for glucose, but glyceraldehyde was markedly inhibitory for the utilization of I4C and tritium. Glyceraldehyde is an unstable compound and is difficult to purify. Commer- cial samples contained up to 30% of impurities by analysis and it is possible that the depressed isotope uptake is due to some contaminant.

9s Glucose Recycling by Hepatocytes

Table 3. Effect of additions on the metabolisni of[3-’H, U-I4C]- and [S-’H, U-‘4C]glucose by hepatocytes f i om fasted rats Means and range of from 4 to 6 rats

Added substrates Glucose synthesis Apparent incorporation of labeled Relative yield 3HOH from [3-3H]glucose glucose in in water ’HOH from [5-’H]glucose

co2 water 3-’H 5-’H

U-I4C 3-’H 5-3H

pmol x 100 mg-’ x h-’ pmol glucose x 100 mg-’ x h-’

None Lactate Pyruvate Fructose Dihydroxyacetone Xylitol Glycerol Glyceraldehyde

-

17-35 14-21 49-82 26 - 42 45 - 54 28 - 39 27 - 39

1.4-2.8 0.3-0.7 0.3-0.6 0.4-1.0 1.4-2.4 0.3-0.6 0.3-0.7 0.5-0.9

4.4- 8.9 1.8- 4.0 0.9- 3.4

5.5-10 1.2- 3.6 1.0- 3.7 1.7- 4.0

3.1- 7.7

5.6-12 100 100 0.73 (0.63-0.78) 5.7- 7.4 43 61 0.53 (0.37-0.63) 2.5- 4.5 37 48 0.55 (0.37-0.76)

7.5-14 104 110 0.71 (0.54-0.87) 5.3- 10 63 44 0.63 (0.40-0.82)

1.8- 5.7 42 38 0.77 (0.63-0.90) 2.0- 5.2 37 42 0.62 (0.49 - 0.76) 5.2- 7.7 35 41 0.58 (0.45-0.72)

Table 4. Effect of glucagon on apparent uptake of 3-’H and 5-’H glucose by hepatocytes from fasted rats Concentration of all substrates 10 mM, values of 2 to 3 experiments. No glucagon = 100%

Additions Effect of glucagon on 3HOH from [3-3H]glucose .~

3HOH from [5-’H]glucose with glucagon

glucose production 3HOH formation from

[3-3H]glucose [5-’H]glucose

%

None Lactate Pyruvate Fructose Sorbitol Dihydroxyacetone Glycerol Xylitol

-

164 134 167” 195 170 116 132

71 61 53 63 51 52 72 81

81 80 52 67 71 66 86 83

0.85 0.43 0.76 0.75 0.65 0.72 0.62 0.95

a The fructose was completely exhausted limiting the effect of stimulation by glucagon.

Lactate, and to a lesser extent pyruvate, depressed detritiation of [3-3H]glucose considerably more than that of [5-3H]glucose, and in the presence of these acids only about half the labeled triose appears to be formed via aldolase cleavage and about half via transaldolase catalyzed exchange.

In Table 4 the effects of glucagon on the detritiation of [3-3H]- and [5-3H]glucose is shown. We con- firm the observations of Clark et al. [9] that glucagon depresses the detritiation of [5-3H]glucose, and a similar depression is apparent with [3-3H]glucose. Again the effect of lactate in depressing preferentially the detritiation of [3-3H]- over [5-3H]glucose is pronounced.

DISCUSSION

Methods for the estimation od recycling are presented in the Appendix. They are restricted in

that they assume the absence of an active pentose cycle. The calculation of phosphorylation of glucose (see Appendix) requires knowledge of the specific activity of glucose-6-P, not available in the present experiments. The apparent rates for the apparent utilization of [2-3H]glucose represent minimal values for the rate of glucokinase. Some rough estimates of the magnitude of underestimate is possible. Preliminary experiments indicate that the specific activities of glucose-6-P from [U-14C]glucose in the presence of lactate in fed rats are about half those of glucose. Since the 3H/14C ratio in glycogen and presumably in glucose-6-P is for the conditions of Table 1, 0.3 to 0.4, the specific activities of tritiated glucose-6-P will be fairly low. Using these values, the rates of glucose- 6-P phosphorylation in the rats fed a high sucrose diet would be, whether lactate is present or not, of the order of 100- 120 pmol x h-’ x gram wet weight-! The rate of glucose 5-phosphatase in the presence of


lactate would be some 20 - 30 % higher and recycling would be very extensive.

In the rat fed a commercial (starch-based) diet the rate of phosphorylation would be of the order of 20 pmol x h-l x 100 mg protein-', and somewhat less in fasted rats. The rate of glucose 6-phosphatase would be up to twice the rate of phosphorylation. In spite of the approximations it is likely that the estimate of the magnitude of recycling is valid for glucose as sole substrate or with lactate. A major conclusion is that recycling is extensive in livers of animals on high carbohydrate diets with elevated glycogen content and that the phosphorylation of glucose is not depressed by gluconeogenesis. From Table 2 it is apparent that the 3H/14C ratio in glucose-6-P depends on the nature of the precursor. When the ratio is high as with fructose, the corrections become rather im- portant and the apparent rate of uptake underestima- tes considerably of the rates of phosphorylation.

Bloxham et al. [20,21] derived methods based on the use of [5-3H]glucose, to measure fructose-6-P phosphorylation and recycling between fructose-6-P and fructose-1 ,6-P2 in muscle. Essentially this method was also extended to apply to hepatic recycling in vivo [22] and in hepatocytes [9]. While the approach may be valid for muscle which lacks glucose 6-phosphatase and transaldolase, it is in our opinion not applicable to liver. They assumed [22] that in vivo, the activity in 3HOH isolated from liver after intraperitoneal injec- tion of [5-3H]glucose, represent the activity liberated by hepatic metabolism. However the distribution of water in the body is so rapid [23] that the activity in liver water originated mainly froin extrahepatic catabolism.

For their calculation in hepatocyte they assumed that detritiation of [5-3H]glucose occurs solely after aldolase cleavage of fructose-1,6 P2. This was questioned by Hue and Hers [lo], who suggested that the major route for detritiation was from glyceraldehyde-P formed by transaldolase exchange between carbons 4, 5 , and 6 of fructose-6-P and glyceraldehyde-3-P. Our observations with hepatocytes (see Results) indicates that up to one half of the detritiation of [5-3H]glucose proceeds via transaldolase exchange (see below).

An essential feature of the calculations of Bloxham et al. [20,21] and Clark et al. [9,22] is the assumption that there is no tritium retention from [5-3H]glucose in fructose-1,6-P2. They assumed that the rates of aldolase and triose phosphate isomerase are much faster than the rate of phosphofructokinase and fructose bisphosphatase so that the tritium from [2-3H]- glyceraldehyde-P (formed by cleavage of [5-3H]fructose-l,6-PZ) would be promptly lost, and the equi- libration between the trioses and fructose-l,6-P2 so extensive that the latter would be essentially tritium-free. This appears unlikely. In liver extracts the

activity of aldolase is of the same order as that of phosphofructokinase [24]. The activity of triose phosphate isomerase in extracts is nearly 1000 times that of phosphofructokinase [24]. However it has been shown that in rabbit liver in vivo the triose phosphates are not equilibrated [25]. In adipose tissue extracts, the rate of triose phosphate isomerase is over 100 times that of phosphofructokinase, but the rate of isomerization was calculated to be 3 to 5 times the rate of glucose phosphorylation [26]. Calculations for liver not reported here in detail, indicate that unless the rates of aldolase and triose phosphate are extremely high, the fructose-1 ,6-P2 would contain considerable tritium from [5-3H]- or [3-3H]glucose, especially if there is an isotope discrimination effect. The occurrence of tritium retention in fructose-1,6-P2 from [2-3H]glyceraldehyde in hepatocytes was shown by us recently [27].

The errors due to the neglect of transaldolase exchange and neglect of tritium retention in fructose- l,6-P2 are in opposite directions and would partially cancel out. As shown in the appendix the calculation of the rate of phosphofructokinase depend to a large extent on the difference in the specific activities of fructose-6-P and fructose-l,6-P2, a term that appears in the denominator of Eqn (4). If the difference is large, the rate of phosphofructokinase and recycling would be substantial and may be under some conditions equal or greater as the recycling between glucose and glucose-6-P.

[3-3H]Glucose Compared with [5-3H/Glucose

Our results show that the detritiation of [S3H]- glucose exceeds that of [3-3H]glucose. If detritiation were solely from triose phosphate formed in the cleavage of fructose-1 ,6-P2, the 3HOH yield from [3-3H]glucose and [5-3H]glucose is likely to be the same. However while transaldolase catalyzed exchange is likely to contribute to the 3HOH yield from [5-3H]glucose, our findings with [3-3H]glucose support the operation of recycling between fructose-6-P and fructose-1,6-P,. However a substantial part of the detritiation of [3-3H]glucose could occur via the pentose cycle. . Gluconeogenesis from lactate depressed by about

half the detritiation from [3-3H]- and [5-3H]glucose but had little effect on that from [2-3H]glucose (see Table 1). Other substrates depressed detritiation of [3-3H]- and [5-3H]glucose to a lesser extent. Di- hydroxyacetone had a small stimulatory effect but glyceraldehyde was rather inhibitory. We can not offer any satisfactory explanation for these findings. They may be related to some effects on the rate of triose phosphate isomerase or other enzymes. Gluca- gon depressed detritiation from both [3-3H]- and


[5-3H]glucose, confirming the observations of Clark et at. [9].

The Metabolism of Glucose by Rat Hepatocytes

The classical concept of liver, based on extensive experimental data, is that it forms glucose when blood concentrations are low, but that it takes up glucose and converts it to glycogen when concentrations are above threshold values of some 6- 8 mM. On the other hand in hepatocytes there is very little or no glucose uptake at physiological concentrations (below 12mM), and glucose synthesis continues even in the presence of high glucose concentrations, provided low concentrations of precursors (lactate, fructose, etc.) are present.

The uptake of tritium-labeled glucose does not conform to Michaelis-Menten kinetics. The kinetics are sigmoid with activation of phosphorylation by increasing glucose concentration with a maximal fractional rate at 15 mM. This may be explained if the detritiation of [2-3H]glucose depends on two opposing processes, the phosphorylation by glucokinase and dephosphorylation by glucose 6-phosphatase. The K, for glucokinase is about 10 mM [29] and at concentrations between 5 and 15 mM the rate will increase moderately with an increase in glucose. The K, for glucose 6-phosphatase is 2-4 mM, and the concentration of glucose-6-P in the cell 0.1 - 0.3 mM [28]. The dephosphorylation would be in this range nearly proportional to glucose-6-P concentrations. The concept that the uptake of glucose is regulated by the opposing action of the enzymes was suggested by Hue and Hers [28], and the major factor regulating net uptake would be the glucose-6-P concentration. This hypothesis has not yet tested in hepatocytes.

Con elusions

Two roles for futile cycling has been proposed, heat generation and metabolic control [I - 61. The theories are discussed in a forthcoming review [29]. Experimental evidence for the support of these theories is scant. Our present work is mainly descriptive rather than analytical. The results show the importance of studies with tritium-labeled glucose, which bring out novel properties of the metabolism of glucose. The methods to measure phosphorylation and recycling are complex and have yet not been rigorously applied. Alternative procedures not based on the isolation of hexose 6-phosphates have been proposed [29] but have yet not been tested. Further work is needed to measure quantitatively the rates of recycling, and to elucidate the regulatory and physiological role of futile cycling.

Fig. 3. Model for the calculation of the rates of recycling. R,, rate of glucose phosphorylation; R,, rate of hydrolysis by glucose 6-phosphatase; R,, rate of phosphorylation of fructose 6-P (phosphofructokinase); R,, rate of hydrolysis by fructose bisphosphatase. (R,, R,, R3 and R, are expressed in units of mass per unit time)

APPENDIX

Calculations of the Rates of Recycling with Tritium-Labeled Glucoses

The model for the calculations is shown in Fig. 3. Glucose with a specific activity GT is phosphorylated with a rate R , and glucose-6-P is hydrolyzed with a rate R,, so that the net rate of glucose production AG = R, - R,. (If there is glycolysis R , > R2.) It is assumed that the hexose phosphates are completely equilibrated. The rate of phosphofructokinase is R, and that of fructose 1,6-bisphosphatase is R,. There may be inflows and outflows from or into glycogen lactate, trioses, etc. which are not specified. It is assumed that the hexose phosphates are at steady state. It is assumed that the size of the glucose pool is large enough and the time period short enough so that the specific activity of glucose is approximately constant.

Recycling between Glucose and Glucose-6-P

When [2-3H]glucose is the substrate the tritium may be liberated in part in the hexose-6-P isomerization [14] and in part in subsequent steps. It is however assumed that water accounts for nearly all the metab- olized tritium and there is little or none in any other products. This holds provided the contribution of the pentose cycle is small. The retention of tritium in hexose 6-phosphates will depend on a number of factors such as the fraction of tritium transferred in the phosphohexose isomerase reaction exchanged with protons [15], the rate of the isomerization, isotope discrimination, and the rate of glucose-6-P phosphatase. The system has been analyzed in detail in adipose tissue [15] and is also discussed in a review [29] and does not concern us here. It has been calculated that under physiological conditions, there will be substantial tritium retention in glucose-6-P, but little in fructose-6-P.

It is convenient to equate the mean specific activity of glucose to 1.00. We designate the relative specific activities of hexose phosphates by terms such as (Glc-6-P), or (Frud-P),, etc. where the subscripts T and C refer to formation from [2-3H]- and [U-'"C]- glucose respectively. d T(yie1d in water) is the apparent uptake of [2-3H]glucose. The rates R , . . . R 4 . . . AG,


and A T are expressed in units of mass per unit time. Then :

d T = R1 - R2 . (GlC-6-P)T. (1)

Substituting R2 by R, + AG and solving for R, we obtain :

If there is no tritium in glucose-6-P, (Glc-6-P), = 0 and the equation simplifies and R, = AT. Since A T and A G can be readily measured, R, and R, and the recycling between glucose and glucose-6-P are obtained. However as shown by the 3H/'4C ratios (see results), the specific activity of glucose-6-P is not zero and may be high under some conditions.

Calculation ofthe Rate ofFructose-6-P Phosphorylation and of Recycling

M. Clark et al. [9] and Bloxham et al. [20] in their calculations assumed that there is no tritium retention in fructose-1,6-P2 from [5-3H]- (or [3-3H]-)glucose. Our results however [27] suggest retention. Below, calculations of recycling between fructose-6-P and fructose-1,6-Pz are presented on this basis.

The equation below applied to the formation of 3HOH from [3-3H]glucose in liver in the absence of glycogen breakdown and in the absence of the pentose cycle.

If A T is the apparent yield of [3-3H]glucose in water, we have, using the same conventions as for Eqn (2):

A T = R3 . (Fru-6-P), - R4 . (Fru-1,6-P2), . (3)

When there is glucose synthesis from three carbon coinpounds A G = R4 - R3 and substituting for R4 in Eqn (3) and solving for the rate of phosphofmcto- kinase, R3, we obtain:

Thus the calculation requires the estimate of the specific activities of fructose-6-P and fructose-1,6-P2. The concentration of the latter in hepatocytes is of the order of 0.01 mM, and estimation of the specific activity offers considerable difficulty. The equation is more complex when there is glycogen breakdown or synthesis (not discussed here). In such a case changes in both glucose and glycogen must be measured.

Supported by Grant AM 12604 from the National Institute of Health and Grant BMS 7422815 for the National Science Founda- tion.

REFERENCES

1.

2.

3 . 4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14. 15. 16. 17. 18.

19.

Newsholme, E. A. & Gevers, W. (1967) Vitamins Hormones,

Newsholme, E. A. & Start, C. (1973) Regulation in Metubolism,

Newsholme, E. A. (1972) Cardiology, 56, 22-34. Clark, M. G., Bloxham, D. P., Holland, P. C. & Lardy, H. A.

(1973) Biochem. J . 134, 589-597. Clark, M. G., Williams, C. H., Pfeiffer, W. F., Bloxham, D. P.,

Holland, P. C., Taylor, C. A. & Lardy, H. A. (1973) Nature (Lond.) 245,99-101.

Williamson, J. R., Jacob, A. & Scholz, R. (1971) Metabolism,

Clark, D. G., Rognstad, R. & Katz, J. (1973) Biochem. Bio-

Rognstad, R., Clark, D. G. & Katz, J. (1973) Biochem. Bio-

Clark, M. G., Kneer, N. M., Bosch, A. L. & Lardy, H. A.

Hue, L. & Hers, H. G. (1974) Biochem. Biophys. Res. Commun.

Clark, D. G., Rognstad, R. & Katz, J. (1972) J . Biol. Chem.

Schmidt, K., Genovese, J. & Katz, J. (1970) Anal. Biochem.

Zahlten, R. M., Kneer, N. M., Stratman, F. W. & Lardy, H. A.

Rose, I. D. &Rose, Z. B. (1969) Compr. Biochem. 17,543- 161. Katz, J. & Rognstad, R. (1969) J . Biol. Chem. 244, 99- 106. De Wulf, H. & Hers, H. G. (1967) Eur. J . Biochem. 2, 50-56. Zalitis, J. & Oliver, I. T. (1967) Biochem. J . 102, 753-759. Rognstad, R., Clark, D. G. 62 Katz, J. (1974) Eur. J . Biochem.

Clark, D. G., Lee, D., Rognstad, R. & Katz, J. (1975) Biochem.

25, 1 - 87.

John Wiley and Sons, New York.

20,13-26.

phys. Res. Commun. 54, 1141 - 1148.

phys. Res. Commun. 54, 1149-1156.

(1974) J . Biol. Chem. 249, 5695-5703.

58,532-539.

249, 2028-2036.

39,170-179.

(1974) Arch. Biochem. Biophys. 33, 186- 190.

47,383-388.

Biophys. Res. Comm. in press. 20. Bloxham, D. P., Clark, M. G., Holland, P. A. & Lardy, H. A.

21. Bloxham, D. P., Clark, M. G., Goidberg, D. M., Holland,

22. Clark, M. G., Bloxham, D. P., Holland, P. C. & Lardy, H. A.

23. Katz, J., Dunn, A,, Chenoweth, M. & Golden, S. (1974) Bio-

24. Knox, W. E. (1972) Enzyme Patterns in Fetal, Adult and Neo-

25. Rose, I. A,, Kellermeyer, R., Stjernholm, R. & Wood, H. G.

26. Katz, J., Landau, B. R. & Bartsch, G. E. (1960) J. Biol. Chem.

27. Rognstad, R., Wals, P. A. & Katz, J. (1975) J . Biol. Chem.

28. Hue, L. & Hers, H. G. (1974) Biochem. Biophys. Res. Commun.

29. Katz, J. & Rognstad, R. (197.5) Current Trends in Cellular

(1973) Biochem. J . 134,581 - 586.

P. C. & Lardy, H. A. (1973) Biochem. J . 134, 586-587.

(1974) J . Biol. Chem. 249, 279-290.

chem. J. 192,171 - 183.

plastic Rut Tissues, S . Karger, Basel, New York.

(1962) J . Biol. Chem. 237, 3325-3331.

241, 727-740.

in press.

58, 540- 548.

Regulation (Horecker and Stadtman, eds) vol. 10, in press.

J. Katz, P. A. Wals, S. Golden, and R. Rognstad, Cedars-Sinai Medical-Research Institute, 4751 Fountain Avenue, Los Angeles, California, U.S.A. 90029

recycling of glucose by rat hepatocytes

Documents