AVOIDING A ‘SUPERFICIAL’ UNDERSTANDING OF THE EXCRETION OF WATER; importance of ‘thought’ experiments.

Our objective is to define how the kidney handles a water load by asking two questions. First, “ What is the actual signal for the release of vasopressin (VP)? ” Our data suggest that the signal is related to the arterial and not the venous plasma Na concentration (P Na ). In more detail, when humans consumed identical water loads after an overnight fast slowly (in 2‐hr) or rapidly (0.5‐hr), there was an equal venous P Na , but a 4‐mM lower arterial P Na 30‐min after rapid water ingestion; only these latter subjects had a large water diuresis (12±1 ml/min). Second, “ How can we estimate of the volume of filtrate delivered the distal nephron to understand what the urine flow rate should be when VP does not act ’? We found a large discrepancy in this estimation using 3 different data sets in rats. First, the (TF/P) Inulin rose 2‐fold between the PCT and the DCT. Second, the medullary interstitial osmolality was ~2500‐mOsm/kg H 2 O in water‐deprived rats. Third, AQP1 was not found in the descending thin limbs (DtL) of superficial nephrons. We conclude that apparent discrepancies in data can be resolved by recognizing heterogeneity of sampling. First, a lower arterial P Na is needed to induce a water diuresis; hence water that was ingested slowly can be retained for future heat dissipation. Second, washout of the medulla can only influence loops of Henle that have water permeability. Although this seems not to apply to superficial nephrons, examining other data such as the (TF/P) Inulin implies that the DtL is permeable to water even if AQP‐1 were truly absent in superficial nephrons. These points indicate that deeper understanding requires integration of all relevant data.