The role of serine as a possible intermediate of the alternative pathway from glucose to glycogen was investigated under basal and insulin-stimulated conditions in 18-day cultured foetal-rat hepatocytes because these cells cannot use pyruvate-derived metabolites [Bismut & Plas (1989) Biochem. J. 263, 889-895]. Incubation of cells with [U-14C]glucose for 24 h led to a release of labelled serine in the medium concomitantly with a net serine production (100 nmol/24 h per culture). The rate of [14C]serine formation (close to 3 nmol/h per culture) indicated that a large part of newly formed serine originated from glucose. When short-term experiments were performed at day 2, glycogen labelling from [U-14C]serine or [U-14C]glycine, which was increased 3-fold by insulin after 2 h, evidenced their participation as glycogenic precursors. When a double-isotope procedure with [U-14C,3-3H]glucose was used, the direct and the alternative pathways from glucose were found to contribute to glycogenesis by 75 and 25% respectively. Cycloserine (18 mM), a transaminase inhibitor, strongly inhibited glycogen labelling from [U-14C] serine while producing a 70% increase in glucose incorporation by the alternative pathway, in both the presence and the absence of insulin. The inhibitor had no effect on the direct pathway from glucose to glycogen. Supplementation with 1 mM-hydroxypyruvate, a serine-derived metabolite, did not affect direct glucose incorporation, whereas the alternative pathway was stimulated whether insulin was present or not. These results indicate that the sequence glucose----serine----glycogen is operative in cultured foetal hepatocytes. The alternative pathway interferes with hydroxypyruvate utilization, and is likely mediated by the serine aminotransferase pathway, independently of the acute glycogenic action of insulin.

Gopher et al [Gopher, A., Vaisman, N., Mandel, H. & Lapidot, A. (1990) Proc. Natl Acad. Sci. USA 87, 5449–5453] recently reported that about 50% of the glucose formed from [U‐13C]fructose infused nasogastrically in children contained 13C3 adjacent to 13C4. Assuming a high isotopic dilution of the triosephosphate pool, the authors concluded that about 50% of the fructose converted to glucose in liver and intestine bypassed the classical aldolase pathway, utilizing a hypothetical direct pathway that would involve the phosphorylation of fructose 1‐phosphate to fructose 1,6‐bisphosphate. The present work was undertaken in order to establish to what extent the conversion of fructose to glucose in the intestine could account for this unexpected isotopic distribution. The technique of everted sleeves was used to define the rate of conversion of [U‐14C]glucose and [U‐14C]fructose in the small intestine of 24‐h‐fasted rabbits. It appeared that, at the low concentration of fructose used by Gopher et al., almost as much fructose was converted to glucose as remained unmodified in the tissue. Fractose uptake was not inhibited by glucose, and the presence of all the necessary enzymes in the tissue indicated that the fructose to glucose conversion occurred by the aldolase pathway. Remarkably, this conversion operated with an isotopic dilution not exceeding 25%, due to the low rate of glucose metabolism and the near absence of gluconeogenesis from lactate. It can, therefore, be postulated that, in the presence of pure [U13C]fructose, the triosephosphate pool is highly enriched in 13C with little dilution by 12C, essentially giving rise to [U‐13C]glucose, as reported by Gopher et al. There is, therefore, no need to postulate the participation of a direct pathway.

The pathways of glycogen synthesis from glucose were studied using double-isotope procedures in 18-day cultured foetal-rat hepatocytes in which glycogenesis is strongly stimulated by insulin. When the medium containing 4 mM-glucose was supplemented with [2-3H,U-14C]glucose or [3-3H,U-14C]glucose, the ratios of 3H/14C in glycogen relative to that in glucose were 0.23 +/- 0.04 (n = 6) and 0.63 +/- 0.09 (n = 8) respectively after 2 h. This indicates that more than 75% of glucose was first metabolized to fructose 6-phosphate, whereas 40% reached the step of the triose phosphates prior to incorporation into glycogen. The stimulatory effect of 10 nM-insulin on glycogenesis (4-fold) was accompanied by a significant increase in the (3H/14C in glycogen)/(3H/14C in glucose) ratio with 3H in the C-2 position (0.29 +/- 0.05, n = 6, P less than 0.001) or in the C-3 position (0.68 +/- 0.09, n = 8, P less than 0.01) of glucose, whereas the effect of a 12 mM-glucose load (3.5-fold) did not alter these ratios. Fructose (4 mM) displaced [U-14C]glucose during labelling of glycogen in the presence and absence of insulin by 50 and 20% respectively, and produced under both conditions a similar increase (45%) in the (3H/14C in glycogen)/(3H/14C in glucose) ratio when 3H was in the C-2 position. 3-Mercaptopicolinate (1 mM), an inhibitor of gluconeogenesis from lactate/pyruvate, further decreased the already poor labelling of glycogen from [U-14C]alanine, whereas it increased both glycogen content and incorporation of label from [U-14C]serine and [U-14C]glucose with no effect on the relative 3H/14C ratios in glycogen and glucose with 3H in the C-3 position of glucose. These results indicate that an alternative pathway in addition to direct glucose incorporation is involved in glycogen synthesis in cultured foetal hepatocytes, but that insulin preferentially favours the classical direct route. The alternative foetal pathway does not require gluconeogenesis from pyruvate-derived metabolites, contrary to the situation in the adult liver.

The coupling of glycolysis to serine and glycine metabolism was studied in fast-growing Zajdela hepatoma cultured cells. During the exponential phase of growth, occurring between 12 and 72 h, cells exhibited a decreased glycogen content together with a high glycolytic activity. Glycogen labelling, evaluated by 1 h-pulse experiments with [U-14C]glucose (5.5 mM), was minimal during the first 48 h and increased 2.5-fold at 72 h and 8-fold at 96 h, at which times it was also stimulated 2-fold by 10 nM insulin. [U-14C]Glucose carbons were incorporated into nucleic acid bases, with maximal incorporation at 72 h, the rate of nucleotide base labelling exceeding that of glycogen during the first 2 days of culture. Incubation of the cells with [U-14C]glucose resulted in the release into the medium of 14C-labelled glycine, the first intermediate formed on the route from serine to DNA. The rate of release per cell decreased as a function of cell growth, concomitantly with an increased rate of glucose carbon incorporation into nucleotide bases. The latter implied the intermediary formation of amino acids since the transaminase inhibitor cycloserine (10 mM), which totally inhibited [14C]glycine release, decreased by 65% nucleotide labelling from [U-14C]glucose. A dose-dependent inhibition by serine of the rate of [U-14C]glucose carbon incorporation into nucleotide bases was observed, which was maximal at 5 mM serine. These metabolic flux measurements indicate that glucose can be used as a precursor of nucleic acid synthesis. These results strongly suggest that this process is to a large extent mediated by a serine/glycine-biosynthesis-mediated pathway, and reinforce the hypothesis that glycolysis contributes to enhancing the provision of precursors required for cell proliferation.