The secretion of proteins from cells presents a fundamental biochemical problem as all protein synthesis occurs in the cytoplasm, for a protein to be secreted it must cross a membrane. Secreted proteins, like all water-soluble proteins, have extensive hydrophilic regions. Membranes, however, have evolved as permeability barriers to such polar molecules. How is this apparent paradox resolved? Secreted proteins do not seem to be special in any way that would imply they can transiently enter the hydrophobic core of the membrane, leading to the suggestion that an aqueous environment is created across the membrane to allow their passage [l]. Although this idea was first put forward almost two decades ago, it has never been demonstrated experimentally.