Resting cells of Staphylococcus aureus displayed a phosphate (Pi) exchange that was induced by growth with glucose 6-phosphate (G6P) or sn-glycerol 3-phosphate (G3P). Pi-loaded membrane vesicles from these cells accumulated 32Pi, 2-deoxyglucose 6-phosphate (2DG6P) or G3P by an electroneutral exchange that required no external source of energy. On the other hand, when vesicles were loaded with morpholinopropane sulfonic acid (MOPS), only transport of 32Pi (and L-histidine) was observed, and in that case transport depended on addition of an oxidizable substrate (DL-lactate). In such MOPS-loaded vesicles, accumulation of the organic phosphates, 2DG6P and G3P, could not be observed until vesicles were preincubated with both Pi and DL-lactate to establish an internal pool of Pi. This trans effect demonstrates that movement of 2DG6P or G3P is based on an antiport (exchange) with internal Pi. Reconstitution of membrane protein allowed a quantitative analysis of Pi-linked exchange. Pi-loaded proteoliposomes and membrane vesicles had comparable activities for the homologous 32Pi: Pi exchange (Kt's of 2.2 and 1.4 mM; Vmax's of 180 and 83 nmol Pi/min per mg protein), indicating that the exchange reaction was recovered intact in the artificial system. Other work showed that heterologous exchange from either G6P- or G3P-grown cells had a preference for 2DG6P (Kt = 27 microM) over G3P (Kt = 1.3 mM) and Pi (Kt = 2.2 mM), suggesting that the same antiporter was induced in both cases. We conclude that 32Pi: Pi exchange exhibited by resting cells reflects operation of an antiporter with high specificity for sugar 6-phosphate.(ABSTRACT TRUNCATED AT 250 WORDS)