Serine is a vital amino acid for cellular metabolism and is synthesized in large amounts (about 1 g/day in the rat) by most tissues of the body (Snell, 1984). The biosynthetic pathway branches from glycolysis at 3-phosphoglycerate, and proceeds via phosphohydroxypyruvate and phosphoserine to form serine (Snell, 1986). Unusually, end-product control o f this biosynthetic pathway by serine is exerted at the final step, that catalysed by phosphoserine phosphatase, which has a control coefficient in normal liver of 0.98 (D. A. Fell & K. Snell, unpublished work). Of the alternative intracellular uses of serine, perhaps the most significant is its role as the major supplier of one-carbon tetrahydrofolate precursors for nucleotide biosynthesis (Rowe et al., 1985). The direction of serine towards nucleotide formation is effected by serine hydroxymethyltransferase, which catalyses the transfer of C-3 of serine to tetrahydrofolate to form 5,10-methylenetetrahydrofolate, and the formation of glycine. 5,l0-Methylenetetrahydrofolate is the direct carbon donor for thymidylate synthesis (and indirectly for purine synthesis) and glycine contributes both carbon and nitrogen for purine synthesis. The stimulation of DNA synthesis in resting lymphocytes by mitogens is a useful model system for studying the control of cell replication and the metabolic pathways associated with DNA precursor synthesis (Ling & Kay, 1975). In the present work, human peripheral blood lymphocytes were isolated in Ficoll-Paque solution (Pharmacia Inc., Hounslow) and were cultured (at about 6 X los cells/ml) at 37°C in 1 ml of RPMI-1640 medium (Gibco Biocult, Uxbridge), containing 20 mM-Hepes buffer, 2 IIIM-Lglutamine, 500 i.u. penicillin and streptomycin/ml, and 10% (w/v) dialysed fetal calf serum. Half the cultures were stimulated with the mitogenic lectin, phytohaemagglutinin, at a final concentration of 5 pg/ml. Non-stimulated cultures served as controls. Cultures were harvested for measurements every 24 h, at which time the remaining cultures were gently aspirated to maintain the cells in suspension. Five hours or two hours before harvesting, cultures were pulsed with 0.7 pCi of ~-[3-'~C]serine (specific activity 56 mCi/ mmol) or 0.5 pCi of ~-[U-'~C]glucose (specific activity 263 mCi/mmol), respectively. Radioactivity incorporated from serine into total nucleic acids was measured as previously described (Eichler et al., 198 1). Radioactivity incorporated from glucose into (serine + glycine) was determined after separation by t.1.c. of a sonicated cell extract with phenolsaturated water/absolute ethanol/water (1 5:4:1) as the mobile phase. Although the solvent system does not clearly separate serine from glycine, this was not considered a disadvantage since the only known metabolic route for glycine formation from glucose is via serine. Thus, the area of silica removed for scintillation counting encompassed both serine and glycine (as localized using amino acids standards run concurrently). At the time of harvesting, some cultures were assayed for serine hydroxymethyltransferase as described by Snell et al. (1987). After mitogenic stimulation of lymphocyte cultures, there was a parallel increase in both serine synthesis from glucose and nucleic acid synthesis from serine (Fig. 1). Incorporation from [3-14C]serine into nucleic acids was accompanied by a parallel increase in serine hydroxymethyltransferase activity (Fig. 1 ) and [6-'H]thymidine incorporation into nucleic acids (results not shown). The measured increases of radiolabel incorporation into nucleic acids and serine hydroxymethyl-