The endothelial pump of the cornea.

still find it difficult to understand why the cornea expends metabolic energy to maintain its status quo. I suppose that one day we may know why the cornea is not a fixed and static structure, but meanwhile we slowly advance towards an understanding of the dynamic balances underpinning the superficial stasis of corneal transparency. The corneal stroma continuously strives to swell by imbibing fluid mainly from the aqueous humor across the endothelium. It does not do so because of the presence of an active "pump" in the endothelium working in the opposite direction, translocating material from the stroma back into the aqueous humor. Corneal water relations are maintained constantly therefore mainly by this "pump-leak" mechanism operating across the endothelium. The question arose: What is pumped across the endothelium? This can be rephrased to: What is the primary metabolic event which regulates the transparency of the cornea? The question was attacked with vigor by, amongst others, David Maurice, Jorge Fischbarg, and myself, all working independently but in a spirit of friendly cooperation. The answer was, I suppose, fairly obvious when it came, but most of us opted for the wrong choice initially. The basic data on the endothelial "pump" came from two types of experiment. In the first, corneal thickness was recorded with the specular microscope. In the second, electrical activity was measured across the de-epithelialized cornea consisting of the stroma, Descemet's membrane, and the endothelium. Everybody agreed on the facts surrounding the "pump." Activity was abolished in the absence of sodium and reduced in the absence of bicarbonate. The presence or absence of chloride had no effect. The polarity of the associated electrical potential was lens-side negative, which was curious—even puzzling. If a single ion is pumped across any epithelial membrane, the direction in which the ion is pumped automatically develops the "sign of the ion." In short, if the endothelium pumped only one ion, that ion had to be negatively charged, but there was no obvious candidate for the role. The modeling of corneal water relations started. My own contribution was to assume that salt was pumped into the aqueous humor.In itself, that act would produce no electrical potential so I invoked a back-flow of salt across the endothelium to fit the ex-