The monoacylglycerol 2-arachidonoylglycerol (2-AG) has been recently suggested as a possible endogenous agonist at cannabinoid receptors both in brain and peripheral tissues. Here we report that a widely used model for neuronal cells, mouse N18TG2 neuroblastoma cells, which contain the CB1 cannabinoid receptor, also biosynthesize, release and degrade 2-AG. Stimulation with ionomycin (1-5 microM) of intact cells prelabelled with [3H]arachidonic acid ([3H]AA) led to the formation of high levels of a radioactive component with the same chromatographic behaviour as synthetic standards of 2-AG in TLC and HPLC analyses. The amounts of this metabolite were negligible in unstimulated cells, and greatly decreased in cells stimulated in the presence of the Ca2+-chelating agent EGTA. The purified component was further characterized as 2-AG by: (1) digestion with Rhizopus arrhizus lipase, which yielded radiolabelled AA; (2) gas chromatographic-MS analyses; and (3) TLC analyses on borate-impregnated plates. Approx. 20% of the 2-AG produced by stimulated cells was found to be released into the incubation medium when this contained 0.1% BSA. Subcellular fractions of N18TG2 cells were shown to contain enzymic activity or activities catalysing the hydrolysis of synthetic [3H]2-AG to [3H]AA. Cell homogenates were also found to convert synthetic [3H]sn-1-acyl-2-arachidonoylglycerols (AcAGs) into [3H]2-AG, suggesting that 2-AG might be derived from AcAG hydrolysis. When compared with ionomycin stimulation, treatment of cells with exogenous phospholipase C, but not with phospholipase D or A2, led to a much higher formation of 2-AG and AcAGs. However, treatment of cells with phospholipase A2 10 min before ionomycin stimulation caused a 2.5-3-fold potentiation of 2-AG and AcAG levels with respect to ionomycin alone, whereas preincubation with the phospholipase C inhibitor neomycin sulphate did not inhibit the effect of ionomycin on 2-AG and AcAG levels. These results suggest that the Ca2+-induced formation of 2-AG proceeds through the intermediacy of AcAGs but not necessarily through phospholipase C activation. By showing for the first time the existence of molecular mechanisms for the inactivation and the Ca2+-dependent biosynthesis and release of 2-AG in neuronal cells, the present paper supports the hypothesis that this cannabimimetic monoacylglycerol might be a physiological neuromodulator.