Because the V ATPase inhibitors that
Because the V-ATPase inhibitors that have been employed in these studies (including bafilomycin and concanamycin), are membrane permeant, they inhibit all the V-ATPases in the cell. This is important since it is possible that intracellular V-ATPases, in addition to those present at the plasma membrane, might be involved in promoting tumor cell invasion. Similarly, knockdown of a3 using siRNA might reduce plasma membrane V-ATPase expression in tumor erk inhibitor either because a3 is itself present in plasma membrane V-ATPase complexes or a3-containing V-ATPases within intracellular compartments (such as secretory vesicles), are involved in the trafficking of V-ATPases to the cell surface. To directly test whether V-ATPases present at the plasma membrane of breast tumor cells are involved in invasion, two approaches were taken to specifically ablate plasma membrane V-ATPase activity . First, MB231 cells were stably transfected with a V5-epitope tagged version of subunit c, part of the integral V0 domain . The V5 epitope tag was located at the C-terminus of subunit c, ensuring that it would be exposed on the outside of the cell for V-ATPases present at the plasma membrane . Subunit c was chosen because of its presence in 9 copies per complex , thus making it likely that nearly all V-ATPase complexes in the cell would contain at least one copy of the V5-tagged c subunit. In fact, V5-tagged complexes could be detected at the plasma membrane of non-permeabilized cells, indicating that the tagged subunit was assembled and trafficked normally . Next, it was shown that addition of an antibody against the V5-tag inhibited V-ATPase dependent proton flux across the plasma membrane of cells expressing the V5-tagged c subunit but not control, untransfected cells . Importantly, addition of the V5 antibody inhibited both invasion and migration of transfected but not untransfected cells . As a second approach to address the role of plasma membrane V-ATPases in breast tumor cell invasion, a membrane impermeant inhibitor (streptavidin-conjugated biotin-bafilomycin) was employed. Consistent with the above results, this inhibitor significantly reduced both invasion and migration of MB231 cells . These results suggest that it is plasma membrane rather than intracellular V-ATPases that play the most important role in breast tumor cell invasion and migration. Our results have led us to hypothesize that breast tumor cells (by an unknown mechanism) up-regulate the a3 isoform, which results in increased trafficking of V-ATPase complexes to the plasma membrane, where they increase tumor cell invasion (Fig. 3) , , . One possible mechanism by which they might achieve this is by increasing the activity of extracellular proteases that participate in invasion. Cathepsins are an important class of proteases that have been shown to be secreted by tumor cells and have been implicated in tumor metastasis , , , . Secreted cathepsins facilitate tumor cell invasion both by directly degrading extracellular matrix and by activating other matrix-degrading proteases (such as MMPs) which do so. Moreover, because cathepsins normally reside in lysosomes, they require an acidic pH both for their activity and for their proteolytic activation , , . Thus, plasma membrane V-ATPases may in part facilitate tumor cell invasion by activating secreted cathepsins involved in this process. How plasma membrane V-ATPases participate in tumor cell migration is less clear, but could involve interactions with cytoskeletal proteins at the leading edge of migrating cells. Plasma membrane V-ATPases also aid in tumor cell survival  by facilitating the secretion of excess metabolic acid generated through the reliance of tumor cells on glycolytic metabolism. It is interesting to speculate that tumor cells may also increase assembly of V-ATPases at the plasma membrane as a means to increase cell surface V-ATPase activity, and this represents an important but unanswered question.