Return of the secretory kidney

Jared J. Grantham, Darren P. Wallace


The evolution of the kidney has had a major role in the emigration of vertebrates from the sea onto dry land. The mammalian kidney has conserved to a remarkable extent many of the molecular and functional elements of primordial apocrine kidneys that regulate fluid balance and eliminate potentially toxic endogenous and xenobiotic molecules in the urine entirely by transepithelial secretion. However, these occult secretory processes in the proximal tubules and collecting ducts of mammalian kidneys have remained underappreciated in the last half of the twentieth century as investigators focused, to a large extent, on the mechanisms of glomerular filtration and tubule sodium chloride and fluid reabsorption. On the basis of evidence reviewed in this paper, we propose that transepithelial salt and fluid secretion mechanisms enable mammalian renal tubules to finely regulate extracellular fluid volume and composition day to day and maintain urine formation during the cessation of glomerular filtration.

  • tubule
  • glomerulus
  • fluid secretion
  • salt secretion
  • renal evolution
  • fluid balance

the liver, pancreas, salivaryand lacrimal glands secrete relatively large amounts of Na+, Cl, K+, and HCO 3 coupled with the isosmotic movement of water (51). Commonly, this secretate is generated in the proximal portions or acini of a branching tubular arcade, and the chemical composition is modified by converging downstream segments before it is finally discharged from the gland. Sweat glands exhibit similar properties and are counted among the organs of secretion. With the exception of Heidenhain (32), Marshall (44), Fine (21), and a few other heretics who have observed net fluid secretion in isolated tubule segments in vitro (7, 24, 64), the mammalian kidney has not been counted among the organs that produce a watery secretion.

The urine that emerges from mammalian kidneys is widely believed to represent fluid that is left over from incomplete tubular reabsorption of glomerular filtrate (2, 17). In the renal field, “secretion” is a term that is usually reserved to describe the deposition into urine of organic anions and cations, largely by proximal tubules, and of potassium, ammonium, and protons by distal tubules and collecting ducts (45, 49, 57). However, recent evidence demonstrating that isolated nonperfused outer and inner medullary collecting ducts (OMCD and IMCD, respectively) dissected from rat kidneys secrete fluid has reopened the possibility that segmental salt and fluid secretion by mammalian renal tubules may contribute to the formation of the final urine (73, 75). In light of this new information, it is not unreasonable to suppose that the secretory transport of organic and inorganic solutes into the proximal tubules (22), coupled with the downstream secretion of salt and fluid into the collecting ducts, contributes to the formation and the modification of urine by mammalian kidneys. In this brief review, we reexamine the evolutionary basis of renal tubule secretory mechanisms that couple solute and fluid transport into the urine formed by mammalian kidneys and how this process may participate in normal and pathological states.



In his brilliant treatise, From Fish to Philosopher, Homer Smith (60) considered how the kidney may have evolved in the sea from early excretory structures. The leading electrolyte constituents in seawater are (in meq/l) chloride (537), sodium (461), magnesium (53), and calcium (10) (35). Smith reasoned that, with free access to both water and salt, primordial kidneys evolved primarily to discharge potentially harmful chemicals that could not be eliminated by simple diffusion through the integument of increasingly complex multicellular animals. These waste products included polar organic compounds that, in early protovertebrates, were secreted into the lumens of simple tubular structures open to the coelom at one end and to the sea at the other. In this context, the secretory transport of solutes, including NaCl and organic molecules, coupled with the osmotic flow of water, also served to propel liquid along the tubule for eventual discharge into the sea. This primitive apocrine kidney appears to represent the framework on which additional tubular excretory processes were added in the course of evolution (Fig. 1).

Fig. 1.

Schematic view of tubule NaCl and fluid secretion over the course of evolution. The protovertebrate tubule provided for the elimination of organic “waste” or xenobiotics through a simple canal mechanism. Urine flow was generated by the influx of liquid along the tubule due to the secretion of organic molecules and salt and, possibly, by the sweeping action of cilia. The glomerular apparatus was added to the mesonephric tubule to facilitate the elimination of relatively large amounts of salt-free water that diffused through the skin or was absorbed from the gut. These tubules greatly increased the reabsorptive mechanisms for conserving NaCl, thereby obfuscating any roles the less powerful secretory processes for organic anion/cation and salt secretion might have had in the generation and modification of urine. The return of some animals to the sea diminished the importance of glomerular filtration for the elimination of water and the conservation of salt, leading to atrophy or complete loss of glomeruli. The metanephric tubule retained the proximal and collecting duct capacities for solute and fluid secretion; however, the relative dominance of glomerular filtration appears to vary depending on whether the animals live in a freshwater aquatic or desert environment. There are no known examples of aglomerular mammals living on land or in the sea. It is important to note that, although the metanephric collecting ducts arise from a different analage (ureteric bud) than either the pronephric or mesonephric tubules, they harbor molecular machinery for the secretion and reabsorption of salt and water.

In representatives of the early life forms available for modern study, orthologs of the brush-border-laden proximal tubules exhibit mechanisms for the transepithelial secretion of organic solutes, including polar anions and cations (43). For example, in insect Malphigian tubules, which have no glomeruli attached to them, organic anions and cations are secreted into the lumens of blind-end tubules against steep concentration gradients. The fluid elaborated by these tubules appears to depend on the contemporaneous secretion of electrolyte and nonelectrolyte solutes. In recent work, it has been observed in Malphigian tubules of Aedes aegypti that hormonally regulated chloride secretion through selective anion channels has a significant role in the elimination of massive amounts of solutes and fluid (6, 47).

Elements of these relatively simple protovertebrate kidneys can still be found in the hagfish and lamprey (60) and are transiently represented in the pronephros of early-stage embryonic renal development in reptiles, birds, and mammals (72).

The Vertebrate Kidney

The upheaval of continents from the sea and the consequent exposure to fresh water provided “the theater of evolution of the early vertebrates” kidney (60). Herein, the simple tubular “kidney” assumed additional roles to preserve the composition of the “internal environment” in this radically different medium through the elimination of excess water and the conservation of salt. To this end, a rudimentary filtering device, the glomerulus, a latecomer in the quiet ritual of urine formation and excretion, was introduced at the blind end of the tubule (Fig. 1). The exuberant filtration of extracellular fluid (ECF) into the tubules by the newly acquired glomeruli solved the problem of getting rid of excess free water, but it created another. Mechanisms for the recapture from tubule fluid of precious filtered solutes such as NaCl were required to preserve homeostasis. Mechanisms for the avid reabsorption of Na+ and Cl had to be upgraded along the tubule segments and added to the tubular secretory mechanisms for the elimination of potentially toxic organic anions and cations. Any contribution of tubular NaCl secretion to urine formation carried from the protovertebrate era was buried in the glutinous reabsorption of filtered solutes imposed by the acquisition of glomeruli. Smith opined (60)

What engineer, wishing to regulate the composition of the internal environment of the body on which the function of every bone, gland, muscle, and nerve depends, would devise a scheme that operated by throwing the whole thing out sixteen times a day and rely on grabbing from it, as it fell to earth, only those precious elements which he wanted to keep?

The bony fishes fully exhibit the mesonephric agenda, where precious solutes in the glomerular filtrate are reclaimed for the ECF and the extraneous water is released to the sea. However, on close inspection, the proximal tubule segments of marine and freshwater fishes have also retained mechanisms for transepithelial net solute secretion that can be coupled to fluid secretion. Beyenbach and colleagues (5, 7, 8, 16) dissected proximal tubules from marine- and freshwater-adapted fish and observed that in some, but not all, nonperfused segments, the lumens opened in the dissecting dish despite the absence of glomeruli. Sustained net secretion of fluid isosmotic with the ambient medium was documented by direct measurement, and cAMP accelerated the rate of net chloride and fluid secretion. These studies proved that mesonephric tubules retained the capacity to secrete solutes and water. It is not a stretch of logic to suppose that under in situ conditions, in which glomerular filtration might be halted, these same tubule segments would generate urine by transepithelial secretion of solutes and water. Indeed, studies of intact Southern Flounder in which glomerular filtration was reduced to zero demonstrated unequivocal secretion of urine containing Mg2+, Na+, Cl, and SO 42 (33, 34). Proximal tubules isolated from frog kidneys have been observed to secrete liquid in response to the addition of phenol red to the external medium (35).

The Metanephric Kidney

As emphasized by Smith (60), the kidney did not finally assume full responsibility for regulating body fluid and salt balance until animals emerged permanently onto dry land (Fig. 1). Amphibians, followed by reptiles, acquired mechanisms for excreting dilute fluid created in a unique segment just beyond the proximal tubule that absorbed salt without water, which eventually became the ascending limb of Henle. In this way, the volume of dilute urine could be hormonally regulated in relation to the availability of water. Reptiles, some living away from a source of water for long periods, reduced the importance of glomerular filtration, depending once again on tubular mechanisms to regulate solute and fluid balance to a significant extent (18). But because the osmolality of urine could not be concentrated above that of the body fluids, severely limiting habitats for survival, additional mechanisms were required to conserve water and to regulate the excretion of NaCl. To this end, the collecting duct system arose from the metanephric blastema, and the loop of Henle arrangement of tubule segments and blood vessels in the medulla were interposed between the proximal and distal tubules. With the creation of the countercurrent mechanism for concentrating solutes in the renal medulla, the mammalian kidney expanded the capability of animals to live away from pools of water. For example, the desert rat concentrates urine to more than 4,000 mosmol/kgH2O, in the face of relatively high rates of glomerular filtration, subsisting only on the water derived from the metabolism of relatively dry food.

An easier solution to mammalian survival under extremely arid conditions might have been to reduce glomerular filtration to rates as some desert reptiles and birds have done (10-12). However, as Smith (60) surmised, “the filtration-reabsorption system (of mammals) is now so firmly established that there is no easy way to overhaul it and to convert it to a purely tubular kidney, as the marine fishes have done.” He became resigned to the filtration-reabsorption view of urine formation in mammals but overlooked the collecting duct system in the regulation of NaCl balance, relegating this segment to a relatively passive role in the reabsorption of water under the control of the antidiuretic hormone (60, 61). However, these terminal segments do much more than reabsorb water. In the last two decades, direct studies have established that the cortical and medullary collecting ducts participate in the regulation of Na+, Cl, K+, NH 4+ , and H+ excretion. Despite this, the possibility that the proximal tubule and collecting duct systems may contribute to the regulation of solute and fluid balance in normal and pathological states by secreting electrolytes and fluid is not included in modern textbooks of renal physiology and nephrology.

Solute and Fluid Secretion in Mammalian Proximal Tubules

The capacity of mammalian proximal tubules to secrete as well as to reabsorb solutes and fluid was discovered inadvertently. My colleagues and I were surprised to find that the inclusion ofp-aminohippurate (PAH) or human uremic serum containing increased levels of hippurate in the external medium bathing an isolated perfused S2 proximal tubule segment dissected from the rabbit kidney caused a sustained reversal in the net transport of fluid from absorption to secretion (28) (Fig.2). Hippurate, a normal aryl anion metabolite, was as effective in causing fluid secretion as PAH, and probenecid, which blocked peritubular hippurate transport into the cells, inhibited fluid secretion. The net flux of hippurate was relatively low compared with net sodium transport in this segment; thus the effect of the organic anion on net fluid transport could only be seen at very low rates of tubule perfusion. A luminal PAH concentration of 40 mM had to be reached in the luminal fluid to override the active reabsorptive transport of sodium (22, 28). The peritubular mechanism for cellular uptake is powerful enough to raise intracellular hippurate to levels many-fold greater than in the external medium, thereby favoring hippurate entry into the lumen. This hyperpolarizes the transepithelial electrical potential, promoting Na+transport through a paracellular pathway. In this way, the net reabsorption of NaCl is converted to net secretion of Na-hippurate, NaCl, and fluid. The mechanism is similar in most respects to other tissues that can either absorb or secrete Na+, owing to changes in the amount of secreted anion (58, 70, 71).

Fig. 2.

Secretion of fluid by isolated mammalian renal tubules. Proximal tubule: a nonperfused rabbit S2 segment was incubated in control medium (top), and then human uremic serum containing hippurate was added for 15 min (bottom). A prominent lumen formed as fluid was secreted secondary to the active transport of hippurates from the serum into the urine (adapted from Ref.28). Inner medullary collecting duct: a nonperfused, rat inner medullary collecting duct (IMCD1) was incubated in control medium (top), and then 8-bromoadenosine 3′,5′-cyclic monophosphate, a permeable derivative of cAMP, was added for several hours (bottom). A prominent lumen formed as fluid was secreted secondary to the active transport of salt and water.

We calculated that as much as 1 liter of secreted fluid [0.5% of glomerular filtration rate (GFR)] could be generated by hippurate transport within a 24-h period (22, 24). During periods of normal glomerular filtration, this secretory contribution of hippurate to urine formation would be masked by the robust reabsorption of 179 liters of Na+-rich glomerular filtrate. Thus it is not difficult to appreciate why fluid secretion driven by organic anion transport would be overlooked by even the best clearance methods for quantifying fractional and absolute levels of tubule fluid absorption and secretion. Indeed, an attempt to detect in humans a contribution of transtubular PAH secretion to urine formation failed because the reabsorption of glomerular filtrate was many times greater than the amount of fluid that could be secreted into the urine coupled to the transport of organic anions (4). To demonstrate fluid secretion coupled to hippurate transport by clearance methods, GFR would have to be reduced to <10% of normal to diminish the obscuring effect of competing NaCl-driven tubule fluid reabsorption, i.e., a situation more closely approximating that of the apocrine protovertebrate kidney. Thus the extent to which organic anion secretion may contribute to the entry of fluid into the lumens of mammalian proximal tubules in situ remains to be determined.

Solute and Fluid Secretion in Mammalian Collecting Ducts

Collecting ducts, the last intrarenal tubular segments through which urine flows, were also the last tubular segments in which transepithelial NaCl transport was rigorously quantified. The potential role of collecting ducts in the regulation of NaCl balance was hotly debated in the 1970s (20, 36, 41, 48, 64, 65, 67, 68). The majority view, which holds to this day, supposes that collecting ducts reabsorb, under the control of aldosterone, an amount of NaCl equivalent to ∼3–5% of the filtered load (45,76-78). Na+ reabsorption, linked to the epithelial sodium channel (ENaC) (54), has been demonstrated in the cortical collecting ducts (CCD) (2, 26,27); however, OMCD and IMCD have not been found in in vitro perfused tubule experiments to transport much NaCl one way or the other, with some notable exceptions. Rocha and Kudo (52) reported that rat IMCD absorbed NaCl under normal perfusion conditions, but this could be reversed to net secretion on the addition of cGMP to the external medium or the addition of atrial natriuretic peptide, which stimulates the intracellular production of cGMP. However, efforts in other laboratories to reproduce these results have not been successful. On the other hand, in support of Rocha and Kudo's findings, cultured rat CCD and IMCD appear to absorb Na+under basal conditions but can be induced to secrete Clon the addition of certain agonists (19, 38-40). Recently, Wall (73) has observed net Cl and fluid secretion in isolated perfused OMCD of the rat, although the effect of cyclic nucleotides on the rate of secretion was not tested.

Sonnenberg and colleagues (62-64) used a retrograde catheterization technique to measure NaCl and fluid transport in rat IMCD in situ and found that massive ECF volume expansion with saline converted net NaCl and fluid absorption to net secretion. Results consistent with bidirectional Na+ transport in the IMCD have been reported by using classic in situ micropuncture methods (3); however, there is a concern that admixture of urine from juxtamedullary long-looped nephrons and superficial nephrons may have complicated the interpretation of these results (20, 36,41). Net NaCl secretion in IMCD is further obfuscated by the relatively low rates of net solute transport, together with the fact that renal pelvic contraction (56) and the concentration of interstitial solutes are disturbed by in situ microcatheterization and micropuncture and in vitro microperfusion methods.

Wallace and colleagues (75) have recently addressed the issue of collecting duct solute and fluid transport using a novel strategy to quantify net absorption and secretion. IMCD were dissected from the renal medullas of rats and maintained in vitro at 37°C for several hours. With the use of methods developed for the study of net fluid secretion in proximal tubules, the lumens of nonperfused collecting ducts were examined after addition to the incubation media of substances that had been shown in a variety of secretory tissues, including human renal cyst epithelial cells, to promote the net secretion of Cl and fluid (Fig. 2). The results were striking and unambiguous. The addition of cAMP to the collecting ducts caused previously collapsed lumens to open widely with fluid, albeit at a relatively low rate. Inclusion of benzamil, a high-affinity amiloride derivative that inhibits ENaC channels in IMCD, potentiated the net secretion of fluid caused by cAMP, indicating that reduction of competing Na+ absorption unmasked a significant degree of solute and fluid secretion. After inhibition with benzamil, a small amount of residual absorption implicated nonspecific solute transport mechanisms in addition to ENaC. The solutes secreted into the lumens were osmometrically active and, more than likely, principally comprised Na+ and Cl, although contributions of K+, NH 4+ , and unspecified osmolytes have not been excluded. Inhibitors of Cl secretion [bumetanide, diphenylamino-2-carboxylic acid (DPC), and DIDS] significantly reduced net fluid secretion promoted by cAMP. In preliminary studies, epinephrine was found to be a potent agonist of Cl-dependent fluid secretion, working through β-adrenergic receptor stimulation of adenylate cyclase (74).

Evidence of cAMP-mediated Cl secretion has been observed in cultured monolayers enriched in IMCD cells obtained from rat and mouse (39, 40, 66). In preliminary studies (74), cultured cell monolayers enriched in human IMCD cells from the initial region of the inner medulla generated an apically negative transepithelial potential difference averaging 5 mV and positive short-circuit current (SCC). This basal SCC was diminished by the apical application of benzamil, implicating ENaC-dependent cation absorption as a contributor to the basal SCC. The addition of cAMP to benzamil-treated membranes strikingly increased positive SCC, consistent with the stimulation of anion transport. Bumetanide, DPC, and DIDS inhibited the cAMP-mediated current, further implicating net secretory Cl transport by these cells. These preliminary studies in cultured human cells support the view that IMCD have the intrinsic capacity to secrete Cl and Na+ and thereby osmotically drive water into the tubule lumen.


Regulation of NaCl and Fluid Balance

Under normal conditions in which there is free access to water and salt, the kidneys of a normal person filter ∼125 ml/min of plasma containing 140 and 105 meq/l (25,200 and 18,900 meq/day) of Na+ and Cl, respectively (69). Approximately 100 meq of Na+ and Cl are excreted daily in the urine. Of the 180 liters of water filtered daily, ∼1.0 liter is excreted in the urine. Thus >99% of the filtered NaCl and 179 liters of water are normally reabsorbed in the steady state. The extent to which relatively small rates of tubular NaCl and water secretion might contribute to this urine formation has been impossible to quantify and, consequently, easy to overlook.

Assuming for the moment that certain renal tubule segments do secrete NaCl under normal conditions, how might this have a role in the formation of the final urine and the regulation of salt and extracellular fluid balance? Salt secretion by tubules distal to segments with high Na+ absorptive capacity would seem to have the greatest impact on the economy of salt and water balance. Were collecting ducts to add NaCl to the tubule fluid, it is likely that this contribution would be reflected in the solute content of the final urine. For the purpose of illustration, an extreme example of this hypothesis is shown in Fig. 3, where it is assumed that all of the filtered NaCl is reabsorbed before the urine enters the medullary collecting ducts. In this case, the NaCl in the final urine would be derived entirely from collecting duct secretion. Alternatively, salt absorptive and secretory processes may be distributed to varying degrees along the collecting duct system from cortex to papilla.

Fig. 3.

Illustration of the potential role of IMCD NaCl secretion in intact kidneys. In the classic view (A), most of the NaCl is reabsorbed in the proximal tubules, loop of Henle, and distal convoluted tubules, leaving a small portion to be reabsorbed by the cortical collecting duct (CCD) system. In an extreme hypothetical alternative (B), Na+ reabsorption (similar to K+ reabsorption) could be nearly complete by the end of the distal tubule or CCD, leaving the outer medullary collecting duct (OMCD) and the IMCD to add NaCl to the final urine. Axial blending of the NaCl absorptive and secretory processes along the CCD, OMCD, and IMCD would accomplish the same outcome to finely regulate NaCl balance in the terminal portions of the kidney tubular system.

How then might we assess the possible contribution of salt secretion to urine formation? Clinicians have observed for decades that an imbalance between the intake and the excretion of salt and fluid often occurs insidiously, although eventually leading to massive edema. For example, the daily retention of salt and water equal to 100 ml of ECF (equivalent to only 0.055% of the GFR) could in 1 mo lead to the accumulation of 3 liters of fluid. In other words, changes in net tubule reabsorption equal to <1% of the filtered NaCl load would be sufficient to cause large changes in ECF volume over extended periods of time. Similarly, a decrease in NaCl secretion equal to <1% of the filtered load could have an effect as important as an increase in the fractional reabsorption of a similar magnitude, and we would be unable to tell one mechanism from the other. Conversely, were renal tubules to secrete each day an additional amount of NaCl equal to 100 ml of glomerular filtrate, neglecting compensatory responses, in 1 mo there would be a decrease in ECF volume of 3 liters and, more than likely, a lowering of the blood pressure. Thus seemingly inconsequential changes in tubule salt secretion could ultimately be reflected by changes in ECF volume manifested as edema or conversely as orthostatic hypotension.

Vasoactive intestinal peptide, a powerful Cl secretogogue in the rat that activates adenylyl cyclase (59), increased the renal excretion of NaCl in isolated perfused rat kidneys without causing changes in renal hemodynamics or GFR (53). This finding is consistent with a contribution of net NaCl secretion to the final urine. Luke (42) reported in 1973 that restricting drinking water for 24 h with a constant intake of salt caused rats to increase the net excretion of NaCl, an effect that was mimicked by the administration of vasopressin to euvolemic animals. The authors suggested that the increased excretion of NaCl, which seemed paradoxical at the time, may reflect a mechanism the kidneys use to maintain the tonicity of ECF during dehydration. In this regard, it is important to note that net NaCl secretion in the IMCD would most likely be downhill, in which case salt excretion could be regulated in part by hormonal effects on the permeability of collecting duct segments to Na+ and Cl (55). It is conceivable that simple dehydration unmasks a cAMP-mediated salt secretory mechanism in rat renal tubules that contributes to the regulation of ECF tonicity, a process in the protovertebrate era that was reinstated several million years ago in Lophius, a fish that unloaded its glomeruli on returning to live in the sea (44,60) (Fig. 1).

Perhaps it is time to reexamine, in those patients with mysterious forms of chronic edema or ECF contraction, the possibility that tubule NaCl secretion might have a role to play. Could the impaired renal tubular functions in some idiopathic edema states, e.g., congestive heart failure and hepatic insufficiency, reflect to some extent a diminution in the tubular secretion of NaCl? And, conversely, could excessive tubule NaCl secretion aggravate the salt wastage of certain renal diseases and orthostatic states associated with ECF contraction? Nephrologists may recall that it was not until indomethacin, a cyclooxygenase inhibitor, was administered to intact animals that roles for endogenous eicosanoids were appreciated as determinants of water and salt balance (1). Furthermore, in this regard, inhibition of prostaglandin synthesis in normal rats enhanced Cl reabsorption in collecting duct segments (37), a possible clue that eicosanoids regulate salt transport in these segments through their capacity to stimulate the formation of cAMP and, thereby, stimulate Cl secretion pathways. As we come to better understand the cellular mechanisms of renal tubule NaCl secretion, it may be possible to find more selective inhibitors or promoters of tubule solute and fluid secretion that will make it possible to examine the potential impact of this occult mechanism on long-term ECF balance.

Potential Role of Tubule NaCl and Fluid Secretion in Acute Renal Failure

The sudden reduction of blood volume through external hemorrhage or plasma sequestration leads to changes in renal hemodynamics that ultimately lead to the renal conservation of ECF. Similar effects on the kidney can be produced by the intravascular infusion of powerful renal vasoconstricting agents, or in some animals such as the diving seal, by a combination of neural and humoral mediated factors. The net effect is to reduce renal blood flow, decrease GFR, and minimize the renal loss of NaCl and water. Were glomerular filtration to cease, tubular lumens would collapse as fluid was completely reabsorbed. However, do the tubule lumens remain completely collapsed under such extreme circumstances? Most human plasma contains sufficient hippurate to promote net fluid secretion in isolated proximal tubules (28,50). A similar secretory mechanism has been suggested as a means to maintain tubular excretory function in lower animals in which glomerular filtration is intermittent (7, 13, 15).

It is conceivable that, under conditions of markedly reduced or complete cessation of glomerular filtration, mammalian proximal tubules could secrete hippurate and similar substances in amounts sufficient to maintain patent tubule lumens and to propel urine through the nephron, albeit at markedly reduced rates of flow (Fig. 4). In the collecting duct segments, the net addition of NaCl and water by secretion could contribute further to urine formation. In this way, elimination of some of the potentially toxic products normally excreted by the kidneys, as well as NaCl and water, would serve a useful survival function, which mirrors to a large extent the formation of urine in aglomerular teleosts.

Fig. 4.

When mammalian nephrons “return to the sea.” Under normal conditions, the reabsorption of glomerular filtrate in conjunction with the secretion of solutes by proximal tubules and collecting ducts contributes to the volume and composition of the final urine. In acute renal failure and chronic renal failure states, glomerular filtration may be strikingly reduced or abolished, leaving only tubular secretory mechanisms in surviving nephrons to form the urine. Under conditions leading to dessication without salt depletion, glomerular filtration may be strikingly reduced to conserve water, and tubule salt secretion may contribute to the avoidance of profound hypernatremia. In all of these pathological states, secretory mechanisms contribute to the elimination of potentially toxic substances (uremic waste products) in a sparingly small volume of liquid, thereby resembling aglomerular animals that live in the sea.

The extent to which proximal tubule fluid secretion may occur in intact kidneys under conditions defined as acute renal failure is unknown. Widely dilated renal tubules, commonly seen in photomicrographs of renal biopsy sections, are consistent with a secretory process. Robust secretion of hippurate in patients destined to recover from acute renal failure has been well established and is a reliable indicator of prognosis (30). Whether the scant amount of urine formed under such circumstances is a product of incomplete reabsorption of markedly reduced glomerular filtrate or a product of tubule solute and fluid secretion is a question worth consideration. It is generally held that the relatively high fractional excretion of Na+ and Cl in urine with a markedly reduced urine-to-plasma creatinine ratio is secondary to reduced tubular reabsorption of scant glomerular filtrate attributable to tubule cell mispolarization of transport proteins and cellular necrosis (46). The possibility that the urine may also reflect the secretion of organic and inorganic solutes and fluid by the proximal tubules and the dilution of this fluid by downstream secretion of NaCl and water by the collecting ducts has not been seriously considered. Perhaps it should. Could it be that the favorable effect of aminophylline, a phosphodiesterase inhibitor, on the outcome of acute renal failure may have as much to do with the stimulation of tubule salt and fluid secretion as with the blockade of adenosine receptors (9,31)?

Role of NaCl and Fluid Secretion in Tubule Cyst Formation and Enlargement

Autosomal dominant polycystic kidney disease (ADPKD) (14) and acquired cystic kidney disease (23) have provided important clues to the existence within tubular epithelial cells of mechanisms for the secretion of solutes and fluid. In ADPKD, cysts develop in proximal, distal, and collecting tubules as a consequence of the mutation of either of two genes, PKD1 or PKD2. The cysts begin as focal outgrowths of individual tubule segments and in a sense represent benign neoplastic tumors that are full of liquid rather than cells. In the early stages of cyst formation, the fluid within them derives from the afferent tubule segment in which the cyst arose. Unreabsorbed glomerular filtrate enters the cysts and collects there as the segment expands, owing to the growth of mural epithelial cells. Interestingly, when the cysts reach a diameter of ∼200 μm, most of them separate from the parent tubule and become isolated sacs of liquid. In this circumstance, net transepithelial fluid secretion is the only mechanism for the sustained addition of liquid to the expanding cyst.

The mechanisms of fluid secretion in human renal cysts of ADPKD patients have recently been elucidated (70, 71). The mural cells appear somewhat less mature that the proximal and distal epithelium from which they derived, and the capacity to absorb solutes and water is greatly reduced. On the other hand, the epithelial cells derived from both proximal and distal tubule elements retain the capacity to secrete NaCl and fluid, as was demonstrated in tubules from normal individuals. This fluid secretion process is regulated by intracellular cAMP, which increases the permeability of cystic fibrosis transmembrane conductance regulator-like Cl channels in apical membranes. This, together with the entry of Clfrom the basolateral side of the cells through a Na+-K+-2Cl cotransporter in conjunction with basolateral K+ channels for recycling this ion, leads to the net secretion of Cl into the urine and an increase in lumen negativity. The negative transepithelial potential increases the movement of peritubular Na+ into the lumen through paracellular pathways.

The amount of fluid secretion into cysts can be increased by substances that activate adenylate cyclase, including arginine vasopressin, prostaglandin E2, secretin, vasoactive intestinal polypeptide, phosphodiesterase inhibitors, and an anonymous neutral lipid that has been detected in cyst fluids of patients with ADPKD (25, 29). Fluid secretion can be inhibited by blocking the Na+ pump with ouabain, inhibiting the Na+-K+-2Cl cotransporter with bumetanide, inhibiting a basolateral K+ channel with glybenclamide, blocking the apical cystic fibrosis transmembrane conductance regulator with DPC, and blocking the action of protein kinase A with H-89 and RpcAMP. Fluid secretion in renal cysts, therefore, resembles in most respects the secretion of NaCl and fluid by mammalian renal collecting ducts.


The mammalian kidney reflects its evolutionary precursors to a remarkable extent. It retains many of the molecular and energy-efficient functional features of simple primordial kidneys that regulate fluid balance and eliminate potentially toxic molecules entirely by secretion. However, the modern kidney must deal with an exuberant glomerulus that was fused to the proximal tubule during the adaptive journey in a freshwater environment, where radical expulsion of free water was necessary for survival. Terrestrial reptiles and desert birds have suppressed the impact of the glomerulus, but it has stubbornly flourished even in mammals that prefer to live in arid climates. In the face of avid solute and fluid reabsorption imposed by glomeruli, occult secretory processes in the proximal and collecting tubules of the mammalian kidney, carried forward during the evolution of vertebrates from the sea to dry land, have persisted as mechanisms for the fine regulation of ECF volume and composition day to day, and possibly for the continuance of a small stream of urine even during the failure of glomerular filtration.


The authors thank Drs. Lawrence Sullivan and Thomas DuBose for helpful discussions.


  • This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grants (P01-DK-53763 and P50-DK-57301 from the Department of Health and Human Services to J. J. Grantham), a National Research Service Award (to D. P. Wallace), and the Polycystic Kidney Foundation.

  • Address for reprint requests and other correspondence: J. J. Grantham, Kidney Institute, Univ. of Kansas Medical Ctr., 3901 Rainbow Blvd., Kansas City, KS 66160 (E-mail:jgrantha{at}


View Abstract