The transport of diadenosine polyphosphates into chromaffin granules from bovine adrenal medulla has been studied by using the radiolabeled substrate [3H]Ap5A and the fluorescent substrate analog di(1,N6-ethenoadenosine)polyphosphate, epsilon-(Ap(n)A) (n=3-5). The vesicular concentration increase was time dependent and the substrates were not metabolized to any extent during the transport experiments. The saturation curve indicates the existence of kinetic and allosteric cooperativity during Ap(n)A (diadenosine polyphosphates) transport and could be the result of the presence of various affinity states of the transporter with K values of 16 +/- 1 microM and 75 +/- 6 microM, and corresponding Hill numbers of 2 and 4, when epsilon-(Ap4A) was the substrate. The saturation studies for [3H]Ap5A were performed in a broader concentration range; in this case a three-step curve was obtained with K values of 16 +/- 2 microM, 125 +/- 9 microM, and 545 +/- 11 microM; the corresponding Hill numbers were 2, 4, and 6. This kinetic behavior can be explained on the basis of a mnemonic model, as already demonstrated for the vesicular transport of ATP. The nonhydrolyzable adenine nucleotide analogs, ATPgammaS and ADPbetaS, inhibited the diadenosine polyphosphate transport at concentrations in the millimolar range. Ap(n)A transport was also inhibited by the P2 receptor antagonist suramin, the mitochondrial ATP/ADP exchange inhibitor atractyloside, the proton translocator FCCP, and N-ethylmaleimide.