This study was designed to simulate purine nucleoside phosphorylase (PNP) deficiency by preincubating with guanosine (Guo) to minimize PNP activity while investigating the metabolism of [14C] deoxyguanosine (dGuo) at physiologic concentrations (10 microM) by unstimulated thymocytes, tonsil-derived T and B lymphocytes, and peripheral blood cells over short time periods. GTP was the principal metabolite formed from dGuo by all cell types with functional PNP and hypoxanthine-guanine phosphoribosyltransferase, confirming formation via degradation to guanine with subsequent salvage by hypoxanthine-guanine phosphoribosyltransferase. Thymocytes also formed a small amount of deoxyguanosine triphosphate (dGTP), presumably through direct phosphorylation by deoxycytidine kinase. Incorporation of dGuo into GTP was effectively inhibited in all instances under PNP deficiency conditions and dGTP levels increased up to 10-fold in thymocytes, but tonsil-derived B or T lymphocytes and unfractionated PBL still accumulated no detectable dGTP. E and platelets formed low amounts of dGTP under these conditions. Preincubation with adenine (50 microM) to reverse any Guo-induced toxicity reduced the incorporation of dGuo into GTP without inhibitor in all cell types with intact adenine phosphoribosyltransferase, but had no effect on dGTP accumulation in thymocytes, with or without inhibitor, thus excluding any indirect formation of dGTP via the de novo route. The rapid metabolism of dGuo to GTP, in the absence of PNP inhibition and subsequent effects of the altered GTP concentrations on cellular metabolism, may account for the differing responses reported by investigators with the use of low dGuo concentrations (enhancing), compared with high (inhibitory), concentrations in mitogen-stimulated lymphocyte studies. The exclusive ability of thymocytes to accumulate significant amounts of dGTP, and inability of B cells to do so, provides a logical explanation for the selective T cell immunodeficiency in PNP deficiency.