5'-Deoxy-5'-methylthioadenosine, a by-product of polyamine synthesis, can support the growth of Raji cells in a methionine-free medium, but not the growth of CCL39 cells, although these cells are also able to incorporate radiolabelled 5'-deoxy-5'-methylthioadenosine (MeSAdo) into methionine, S-adenosyl-L-methionine (AdoMet) and proteins [Christa, Kersual, Augé & Pérignon (1986) Biochem. Biophys. Res. Commun. 135, 131-138]. We first tested the hypothesis of a toxic effect of MeSAdo in the conditions of growth experiments: we could not demonstrate any toxic effect of MeSAdo on the synthesis of macromolecules, nor any toxicity mediated by polyamines or pyrimidine starvation, and we found that the growth of CCL39 cells was strictly dependent on the supply of exogenous methionine. We then tried to determine whether the ability of CCL39 cells to metabolize MeSAdo to methionine and AdoMet was modulated by the proliferation state of CCL39 cells, which is dependent on the supply of exogenous methionine. Studies of the incorporation of radiolabelled MeSAdo show that: (i) the total synthesis of methionine from MeSAdo is twice as high in subconfluent cells (grown in 100 microM-methionine) as in resting cells (cultured in 0 microM-methionine); (ii) the incorporation into proteins does not parallel the total protein synthesis, and the methionine derived from MeSAdo mostly flows out of the cell; (iii) addition of methionine to resting cells immediately leads to a transient and marked increase in metabolism of MeSAdo to AdoMet, presumably reflecting the rapid replenishment of the AdoMet pool of the cells. Taken together, these results suggest that the methionine derived from MeSAdo is preferentially used to synthesize AdoMet rather than proteins, and that this synthesis of AdoMet depends on the ability of the CCL39 cells to grow, and hence on the supply of exogenous methionine. It is proposed that, in CCL39 cells, the metabolic pathway leading from MeSAdo (a by-product of polyamine synthesis) to methionine and to AdoMet (a precursor of polyamine synthesis) is part of a metabolic cycle the activity of which depends, like polyamine synthesis itself, on cell proliferation.