pH of TGN and recycling endosomes of H+/K+-ATPase-transfected HEK-293 cells: implications for pH regulation in the secretory pathway.

The influences of the gastric H+/K+ pump on organelle pH during trafficking to and from the plasma membrane were investigated using HEK-293 cells stably expressing the alpha- and beta-subunits of human H+/K+-ATPase (H+/K+-alpha,beta cells). The pH values of trans-Golgi network (pHTGN) and recycling endosomes (pHRE) were measured by transfecting H+/K+-alpha,beta cells with the pH-sensitive GFP pHluorin fused to targeting sequences of either TGN38 or synaptobrevin, respectively. Immunofluorescence showed that H+/K+-ATPase was present in the plasma membrane, TGN, and RE. The pHTGN was similar in both H+/K+-alpha,beta cells (pHTGN 6.36) and vector-transfected ("mock") cells (pHTGN 6.34); pHRE was also similar in H+/K+-alpha,beta (pHRE 6.40) and mock cells (pHRE 6.37). SCH28080 (inhibits H+/K+-ATPase) caused TGN to alkalinize by 0.12 pH units; subsequent addition of bafilomycin (inhibits H+ v-ATPase) caused TGN to alkalinize from pH 6.4 up to a new steady-state pHTGN of 7.0-7.5, close to pHcytosol. Similar results were observed in RE. Thus H+/K+-ATPases that trafficked to the plasma membrane were active but had small effects to acidify the TGN and RE compared with H+ v-ATPase. Mathematical modeling predicted a large number of H+ v-ATPases (8000) active in the TGN to balance a large, passive H+ leak (with PH approximately 10-3 cm/s) via unidentified pathways out of the TGN. We propose that in the presence of this effective, though inefficient, buffer system in the Golgi and TGN, H+/K+-ATPases (estimated to be approximately 4000 active in the TGN) and other transporters have little effect on luminal pH as they traffic to the plasma membrane.