The thick ascending limb and water channels: half-full or half-empty.

THE THICK ASCENDING LIMB of the loop of Henle plays a central role in urine concentration and dilution (reviewed in Ref. 7). The thick ascending limb actively reabsorbs NaCl but has an extremely low transepithelial osmotic water permeability, even in the presence of vasopressin (6). This combination of NaCl reabsorption without water reabsorption serves two vital functions: it provides NaCl to increase the osmolality of the medullary interstitium, tubules, vasculature, and collecting ducts; and it dilutes the luminal fluid within the thick ascending limb. During diuretic conditions, vasopressin levels are low, the collecting duct is virtually water impermeable, and the dilute luminal fluid is excreted as dilute urine. In contrast, during antidiuretic conditions when vasopressin levels are high, the collecting duct becomes highly water permeable, water is reabsorbed osmotically into the hypertonic medullary interstitium, and concentrated urine is excreted. Thus, active NaCl reabsorption without water reabsorption in the thick ascending limb sets up conditions permitting the production of either concentrated or dilute urine, depending on the presence or absence of vasopressin. While the thick ascending limb does not support transepithelial water reabsorption, the basolateral membrane is exposed to changes in interstitial osmolality. The volume of thick ascending limb cells changes in response to changes in interstitial osmolality (2– 4). In an issue of the American Journal of Physiology-Renal Physiology, Cabral and Herrera (1) provide an elegant explanation for this apparent contradiction between the lack of transepithelial water reabsorption and the need for cell volume regulation. They propose that a water channel (or aquaporin) must be present in the basolateral plasma membrane to explain the ability of the cells to volume regulate, but not in the apical membrane to explain the lack of transepithelial water transport. Cabral and Herrera (1) provide molecular and functional data establishing the presence of aquaporin-1 (AQP1) in the basolateral plasma membrane and in the intracellular compartment, but not in the apical membrane. An important aspect of this paper is that it is the first demonstration of AQP1 expression in the thick ascending limb. This is particularly significant since previous studies reported that no aquaporins are expressed in the thick ascending limb (reviewed in Ref. 5). The lack of an aquaporin in the thick ascending limb was not questioned since there is no transepithelial water reabsorption in this segment. However, Cabral and Herrera (1) questioned this conventional wisdom. They reasoned that cell volume regulation does require the presence of an aquaporin, and then went on to show that AQP1 is expressed in the basolateral membrane. Cabral and Herrera (1) went to great lengths to prove that AQP1 is expressed in the basolateral membrane. In particular, their use of isolated thick ascending limbs was critical to their success. AQP1 is expressed in other structures in the cortex and outer medulla, including the proximal tubules, descending limbs, and vasa recta. Thus, its expression may have been missed when studies were performed using whole cortical or outer medullary lysates, or by the use of histochemistry of kidney tissue. In their study, Cabral and Herrera (1) begin to address the physiological importance of AQP1 in the basolateral membrane of the thick ascending limb. Their study of AQP1 knockout mice shows that about half of the basolateral membrane water permeability is mediated by AQP1. Their study did not elucidate the mechanism for the other 50%. One might expect that if AQP1 in the basolateral membrane is essential for cell volume regulation by thick ascending limb cells, then its loss could lead to osmotic damage. However, the AQP1 knockout mouse is a global knockout and has a urine-concentrating defect. Thus, osmotic shifts in the outer medulla will be reduced, mitigating the loss of AQP1 from the basolateral membrane, and hence any osmotic damage. Cabral and Herrera (1) raise a number of intriguing questions that will need to be addressed in future studies, including the mechanism for the 50% of basolateral membrane water flux that is not mediated by AQP1. Their study emphasizes the importance of using kidney tubules from animals and of questioning scientific dogma when it does not explain the physiology. Their study is important as it shows, for the first time, that AQP1 is expressed in the basolateral membrane of the thick ascending limb, but not in the apical membrane. Thus, in terms of AQP1, the thick ascending limb is either half-full (basolateral membrane) or half-empty (apical membrane). Both the half-full and halfempty portions serve key but different physiological roles in the thick ascending limb.