Disorders of Water Homeostasis

Disorders of Water Homeostasis

Tomas Berl

Robert W. Schrier

Historical and Evolutionary Aspects of Renal Concentrating and Diluting Processes

In From Fish to Philosopher, Smith (1) suggested that the concentrating capacity of the mammalian kidney may have played an important role in the evolution of various biologic species, including Homo sapiens. He suggested that the earliest protovertebrates resided in a saltwater environment that had a composition similar to their own extracellular fluid (ECF); therefore, these species could ingest freely from the surrounding sea without greatly disturbing the composition of their own milieu interieur. However, when these early vertebrates migrated into freshwater streams, the evolution of a relatively water-impermeable integument was mandatory to avoid fatal dilution from their hyposmotic, freshwater environment. Thus a vascular tuft—which we now call the glomerulus—developed, enabling the fish to filter the excess fluid from their blood.

The proximal tubule, which reabsorbed isotonic fluid, evolved in response to the need for salt preservation. However, this did not allow the excretion of hypotonic urine, which is critical for the survival of organisms ingesting hypotonic fluid from their freshwater environment. This need was met by the development of the distal tubule, which could dilute tubular fluid and ultimately urine. This dilution is accomplished by reabsorption of salt without water, because the distal tubular epithelium was relatively impermeable to water. The fish then could excrete the excess solute-free water they had obtained from their freshwater environment while concomitantly conserving their body salts.

Vertebrates began to reside on dry land several million years later. The problem of salt conservation persisted in this terrestrial environment, but the excretion of large volumes of dilute fluid was no longer necessary. Rather, conservation of fluid became of primary importance in the new arid environment. The kidneys of reptiles, birds, and mammals, however, had glomeruli, which filtered large amounts of fluid and salt, even though excretion of only minute amounts of these substances was needed to maintain daily balance. In reptiles and birds, the kidneys responded to this challenge by a decrease in the number of capillary loops in their glomerular tufts. Aglomerular kidneys even evolved in some fish, such as the sea horse and pipefish, which may have been the first vertebrates to return to the sea. Tubular secretory systems evolved in these nephrons to allow elimination of nitrogenous wastes without the need for large volumes of filtered fluid. Also, a relatively insoluble nitrogenous end product, uric acid, was produced that could be excreted in supersaturated solutions with minimal water loss.

The high-pressure glomerular filters were maintained in mammals; however, the countercurrent mechanism developed for concentrating urine. Mammals, along with birds, are unique among vertebrates in possessing loops of Henle and in their ability to compensate for water deficits by elaborating urine more concentrated than blood.

Countercurrent Concentrating Mechanism

By analogy with heat exchangers, the functional significance of the loops of Henle was proposed when Kuhn and Ryffel of the physical chemistry department at the University of Basel, Switzerland, originated the concept of the countercurrent multiplier system for urine concentration in 1942 (2). The hypothesis states that a small difference in osmotic concentration (single effect, or einzeln Effekt) at any point between fluid flowing in opposite directions in two parallel tubes connected in hairpin manner can be multiplied many times along the length of the tubes. In the kidney, a small, 200 mOsm gradient results in a large osmolar concentration difference between the corticomedullary junction and the hairpin loop at the tip of the papilla. Since then, although numerous experiments have confirmed the overall operation of a countercurrent multiplier in the kidney, with the thick ascending limb of Henle as the water-impermeable site of active solute reabsorption (3), the precise mechanism that culminates in the generation of the medullary osmotic gradient is not fully understood (4).


As disturbances in the capacity of the kidney to concentrate and dilute the urine are central to the pathogenesis of disorders of water balance, we will briefly review the components of the diluting and concentrating process in the mammalian kidney. These are depicted in Figure 1-1A and B, respectively (5). It is important
to emphasize that many of these processes are the same whether the final excreted urine is hypotonic or hypertonic to plasma.

Figure 1-1 (A) Urinary dilution mechanisms. Normal determinants of urinary dilution and disorders causing hyponatremia. (B) Urinary concentrating mechanisms. Determinants of normal urinary concentrating mechanism and disorders causing hypernatremia. (From Berl T, Parikh C. Disorders of water metabolism. In: London MI, ed. Comprehensive Clinical Nephrology. 5th ed. Philadelphia: Saunders; 2014:94-110.)

Figure 1-1 (continued)

Glomerular Filtration Rate and Proximal Tubular Reabsorption

The rates of glomerular filtration and proximal tubular reabsorption are important primarily in determining the rate of sodium and water delivery to the more distal portions of the nephron, where the renal concentrating and diluting mechanisms are operative. Fluid reabsorption in the proximal tubule is isosmotic; therefore, tubular fluid is neither concentrated nor diluted in the proximal portion of the nephron. Rather, after approximately 70% of glomerular filtrate is reabsorbed in the proximal tubules, the remaining 30% of fluid entering the loop of Henle is still isotonic to plasma. The reabsorption of sodium chloride is primarily driven by the Na/H 3 transporter whereas the isotonic removal of water is facilitated by the robust expression of the water channel aquaporin 1 (AQP1), depicted in Figure 1-2. A decrease in glomerular filtration rate (GFR) or an increase in proximal tubular reabsorption, or both, may diminish the amount of fluid delivered to the distal nephron and thus limit the renal capacity to excrete water. Similarly, a diminished GFR and increased proximal tubular reabsorption may limit the delivery of sodium chloride to the ascending limb, where the tubular transport of these ions without water initiates the formation of the hypertonic medullary interstitium. With diminished delivery of sodium chloride to the ascending limb, the resultant lowering of medullary hypertonicity impairs maximal renal concentrating capacity.

Figure 1-2 Schematic representation of key elements of the kidney tubule that mediate water reabsorption. Direct mediators are those involved in sodium (depicted as red circles), urea (green circles), and water (blue circles) transport. PCT, proximal convoluted tubule; CTAL, cortical thick ascending limb; MTAL, medullary thick ascending limb; DCT, distal convoluted tubule; CCD, cortical collecting duct; OMCD, outer medullary collecting duct; IMCD, inner medullary collecting duct; AQP, aquaporin; NHE, Na+/H+ exchanger; Na/Glucose, sodium-glucose cotransporter; NBC, sodium bicarbonate cotransporter; NaK, Na+/K+ ATPase; NKCC2, Na+/2CI/K+ cotransporter; ENaC, epithelial sodium channel; UT-A, urea transporter A isoform. (From Hasler U, Leroy V, Martin PY, et al. Aquaporin-2 abundance in the renal collecting duct: new insights from cultured cell models. Am J Physiol Renal Physiol. 2009;(297:1), with permission.)

Descending and Ascending Limbs of the Loops of Henle, Distal Tubule, and Collecting Ducts

Because the urine that emerges from the proximal tubule is isosmotic, the first nephron segment actually involved in urinary concentration is the descending limb of Henle’s loop. There are two types of descending limbs. The short loops originate in superficial and mid cortical glomeruli and turn in the outer medulla. The long loops originate in deep cortical and juxtamedullary glomeruli and penetrate variable distances into the inner medulla. Short and long descending limbs are anatomically distinct; the long limbs in particular display considerable interspecies variability (6). Interestingly, no correlation is apparent between a species’ maximal concentrating ability and the ratio of short and long loops. In fact, in rodents with highest urinary concentrations, the number of short loops is considerably greater than the number of long loops. Approximately 15% of nephrons possess long loops in the human kidney; the other 85% of nephrons have short loops. The descending thin limb is very water permeable as it also has, in its first portion, abundant expression of AQP1 (7). Thus, tubular fluid is concentrated as it descends primarily, but probably not exclusively, by the extraction of water.

Somewhat proximal to the hairpin turn, there is a transition from the descending limb to the ascending thin limb of Henle’s loop. This segment, as well as the remainder of ascending limb, is water impermeable, As will be further discussed, the nature and particular site at which the movement of solutes (urea and NaCl) occur has not been fully defined. Active sodium transport has not been demonstrated convincingly, and this segment’s morphologic appearance with few mitochondria does not suggest active metabolic work.

The thick ascending limb of Henle’s loop appears both structurally and functionally distinct from its thin counterpart. The epithelium is remarkably uniform among species with tall, heavily interdigitating cells with large mitochondria. The observation that fluid emerges into the early distal tubule hypotonic (about 100 mOsm/kg H2O) supports the view that active sodium chloride transport out of this water-impermeable segment provides the single effect required for the operation of the countercurrent multiplier. The primary
mechanism of chloride absorption in the thick ascending limb is mediated by an electroneutral sodium, potassium, and chloride (Na+:K+:2Cl) cotransport (Fig. 1-2).

The distal convoluted tubule is the segment between the macula densa and collecting ducts. This is a morphologically heterogenous segment (6) that is also water impermeable and unresponsive to vasopressin. The collecting ducts are formed in the cortex by the confluence of several distal tubules. They descend through the cortex and outer medulla individually, but successively fuse together on entering the inner medulla. In humans, a terminal inner medullary collecting duct draws from as many as 7,800 nephrons. The collecting ducts possess vasopressin-sensitive adenylate cyclase in all species studied; they are virtually impermeable to water in the absence of the hormone. The vasopressin-sensitive water channel AQP2 mediates water reabsorption in this segment of the nephron in concert with AQP3 and AQP4 (7, 8). The collecting duct in its cortical and medullary segments is also impermeable to urea, but in response to the vasopressin-sensitive urea transporter UT 1, the inner medullary collecting duct is rendered urea permeable (9).

Kokko and Rector, as reviewed by Sands (10), have proposed a model of urinary concentration that is in concert with the anatomic features and the permeability characteristics of the various segments of the system, while limiting the active transport of solute to the thick portion of the ascending limb of Henle in the outer medulla. The components of the mechanism, as depicted in Figure 1-3A, are as follows:

  • The water-impermeable thick ascending limb of Henle’s loop actively cotransports sodium, chloride, and potassium, thereby increasing the tonicity of the surrounding interstitium and delivering hypotonic fluid to the distal tubule. Urea is poorly reabsorbed and therefore retained in the tubule.

  • Under the influence of vasopressin in the cortical and outer medullary collecting ducts, tubular fluid equilibrates with the isotonic and hypertonic interstitium, respectively. Low urea permeability in this portion of the nephron allows its concentration to further increase.

  • In the presence of vasopressin, the inner medullary collecting duct is rendered more permeable to urea. Therefore, in this segment of nephron, in addition to water reabsorption, urea is reabsorbed as it diffuses passively along its concentration gradient into the interstitium, where it constitutes a significant component of the medullary interstitial tonicity.

  • The resulting increase in interstitial tonicity creates the osmotic gradient that abstracts water from a highly water permeable and solute impermeable descending limb of Henle’s loop. This process elevates the concentration of sodium chloride in the tubular fluid. When tubular fluid arrives at the bend of the loop, its tonicity is the same as that of the surrounding interstitium. However, the sodium chloride concentration of the tubular fluid is higher and the urea concentration lower than that of the interstitium.

  • Tubular fluid then enters the thin ascending limb, which is more permeable to sodium than urea. The sodium gradient provides for passive removal of sodium chloride from this segment into the interstitium.

To prevent urea removal from the inner medulla to the cortex, the ascending and descending vasa recta act as a countercurrent exchanger and “trap” urea in the inner medulla. The ascending vasa recta also may deposit urea into adjacent descending thin limbs of a short loop of Henle, thereby recycling it to the inner medullary collecting tubule. The descending limbs of short loops do not enter the inner medulla; thus, the addition of urea to these loops does not interfere with the removal of water from the descending thin limb in the inner medulla, a step that is so crucial to the concentrating process.

This passive model of urinary concentration has a number of attractive features, and many of its aspects have been experimentally supported (10). However the requisite difference in sodium and urea permeabilities that is needed for this model to operate passively (point 5 above) have not been met, as mathematical models that employ available permeabilities fail to generate the desired osmotic gradients. Thus, alternatives have been tested. Among these, a three-dimensional reconstruction of the components of the rat inner medulla coupled with the development of mathematical models has emerged with a variation of the previous model, designated as the solute mixing passive model, depicted in Figure 1-3B (4). This model recognizes that the water permeability of the descending limb does not extend into the inner medulla as AQP1 is absent in the lower half of this limb of Henle’s loop. The high urea permeability and the passive exit of sodium from tubular fluid occur before the bend of the loop, equally in the descending as well as ascending thin loops rather than exclusively in the latter. Although this model predicts the generation of concentrated urine, it does not produce one that is maximally concentrated (4).

What remains widely agreed upon is that the single effect in the ascending limb of Henle, so critical to the operation of the countercurrent system and urinary concentration, also serves to dilute the urine. In the absence of vasopressin, and thus with water impermeability of the collecting ducts, the continued reabsorption of solute in the remainder of the distal nephron results in a maximally dilute urine (50 mOsm/kg). Thus, it

should be apparent that impairment of sodium, chloride, and potassium cotransport in the ascending limb of the loop of Henle will limit the renal capacity both to concentrate and to dilute the urine.

Figure 1-3 (A) Schematic representation of the passive urinary concentrating mechanism. Both the thin ascending limb in the inner medulla and the thick ascending limb in the outer medulla, as well as the first part of the distal tubule, are impermeable to water, as indicated by the thickened lining. In the thick ascending limb, active sodium, chloride, and potassium cotransport renders the tubule fluid dilute and the outer medullary interstitium hyperosmotic (1). In the last part of the distal tubule and the collecting tubule in the cortex and outer medulla, water is reabsorbed down its osmotic gradient (2), increasing the concentration of urea that remains behind. In the inner medulla, both water and urea are reabsorbed from the collecting duct (3). Some urea reenters the loop of Henle (data not shown). This medullary recycling of urea, in addition to trapping of urea by countercurrent exchange in the vasa recta (data not shown), causes urea to accumulate in large quantities in the medullary interstitium (indicated by UREA), where it osmotically extracts water from the descending limb (4) and thereby concentrates sodium chloride in descending limb fluid. When the fluid rich in sodium chloride enters the sodium chloride-permeable (but water-impermeable) thin ascending limb, sodium chloride moves passively down its concentration gradient (5), rendering the tubule fluid relatively hyposmotic to the surrounding interstitium. (From Jamison RL, Maffly RH. The urinary concentrating mechanism. N Engl J Med. 1976;295(19):1059, Copyright © 2017 Massachusetts Medical Society. Reprinted with permission from Massachusetts Medical Society.) (B) Illustrating the unique permeabilities and solute fluxes of current solute separation, solute mixing passive model for concentrating urine. Thick tubule border indicates AQP1-null, water-impermeable segment of DTL as well as water-impermeable ATL and TAL. The AQP1-null segment of the DTL is essentially impermeable to inorganic solutes and water. In this model, both the ATLs and the DTLs (including the AQP1-null segment) are highly permeable to urea. In contrast to the original passive model, passive NaCl reabsorption without water begins with the prebend segment and is most significant around the loop bend. Also, in contrast to previous models, urea moves passively into the entire DTL and early ATL, but as this urea-rich fluid further ascends in the ATL, it reaches regions of lower interstitial urea concentration and diffuses out of the ATL again. Thus, the loops act as countercurrent exchangers for urea. (Republished with permission of American Society of Nephrology, from Dantzler W. et al Urine-concentrating mechanism in the inner medulla: function of the thin limbs of the loops of Henle. Clin J Am Soc Nephrol. 2014;9(10):1781-1789; permission conveyed through Copyright Clearance Center, Inc.)

Medullary Blood Flow

Medullary blood flow, whose rate can be regulated independently of whole kidney blood flow, may also affect the renal capacity both to concentrate and to dilute the urine, because the preservation of the medullary hypertonicity in the interstitium is dependent on the countercurrent exchange mechanism in the vasa recta. Although medullary blood flow constitutes only 5% to 10% of total renal blood flow, this flow is still several times more rapid than the tubular flow. The vasa recta possess AQP1 and serve as a countercurrent exchanger that permits the preservation of interstitial tonicity. Blood that enters the descending vasa recta becomes increasingly concentrated as water diffuses out of and solutes diffuse into this portion of the nephron. The hairpin configuration of the vasa recta, however, does not allow the solute-rich blood to leave the medulla. In the ascending portion of the vasa recta, water diffuses into the vasa recta and solute moves out, thus maintaining interstitial hypertonicity. Even with an intact countercurrent exchange system in the vasa recta, circumstances that increase medullary blood flow may “wash out” the medullary concentration gradient and thereby diminish renal concentrating capacity. Moreover, even in the absence of vasopressin, the collecting duct is not completely water impermeable; therefore, a further decrease in the hypertonic medullary interstitium during an increase in medullary blood flow may decrease the vasopressin-independent osmotic water movement from the collecting duct and thereby increase water excretion.

Distal Solute Load

The rate of solute delivery to the collecting duct is a known determinant of renal concentrating capacity. As depicted in Figure 1-4, in spite of maximal levels of vasopressin, urinary osmolality in normal humans progressively diminishes as solute excretion increases. At high rates of solute excretion, urinary osmolality may reach isotonicity in humans, even though supraphysiologic doses of vasopressin are infused. An
increase in solute excretion even may be associated with hypotonic urine with the infusion of submaximal doses of vasopressin in patients with pituitary diabetes insipidus. At least two factors may be responsible for this effect of solute excretion on renal concentrating capacity. First, a solute diuresis generally is associated with an increase in medullary blood flow, which could lower the medullary solute concentration profile. Second, the rapid rate of tubular flow through the medullary collecting duct could shorten contact time sufficiently so that complete osmotic equilibrium of fluid would not be allowed between the collecting duct and medullary interstitium, even though vasopressin had made the collecting duct membrane maximally permeable to water.

Figure 1-4 Effect of solute excretion on renal concentration and diluting mechanisms. The submaximal response to antidiuretic hormone (ADH) may result from the presence of submaximal amounts of ADH or the diminished response of the collecting duct to maximal amounts of ADH. (Reproduced from de Wardener HE, del Greco F. Influence of solute excretion rate on production of hypotonic urine in man. Clin Sci. 14(4):715-723. © 1955, the Biochemical Society.)

Antidiuretic Hormone

The renal concentrating and diluting processes are ultimately, and most importantly, dependent on the presence or absence respectively of arginine vasopressin (AVP) to modulate the water permeability of the collecting duct. AVP, a cyclic hexapeptide (mol wt 1,099) with a tail of three amino acids, is the antidiuretic hormone (ADH) in humans (Fig. 1-5). The presence of a basic amino acid (arginine or lysine) in the middle of the intact hormone at position 8 is crucial for antidiuresis, as is the asparagine at position 5. AVP is synthesized in the supraoptic and paraventricular magnocellular nuclei in the hypothalamus. In these nuclei, a biologically inactive macromolecule is cleaved into the smaller, biologically active AVP. Both oxytocin and AVP are encoded in human chromosome 20 in close proximity to each other, depicted in Figure 1-6. The preprohormone gene is approximately 2,000 base pairs in length and comprises three exons (Fig. 1-6). AVP is encoded in the first exon following a signal peptide. Although spanning all three exons, the binding protein neurophysin is encoded primarily in exon B and the terminal glycoprotein in exon C. The promoter has cis-acting elements, including a glucocorticoid response element, a cyclic adenosine monophosphate (cAMP) response element, and four AP-2 binding sites (11). The precursor prohormone, called propressophysin, is cleaved by removal
of the signal peptide after translation. Vasopressin, with its binding protein neurophysin II, and the glycoprotein are transported in neurosecretory granules down the axons and stored in nerve terminals in the pars nervosa. There is no known physiologic role of the neurophysins, but they neutralize the negative charge of vasopressin. The release of stored peptide hormone and its neurophysin into the systemic or hypophyseal portal circulation occurs by an exocytosis. With increased plasma osmolality, electrical impulses travel along the axons and depolarize the membrane of the terminal axonal bulbs. The membrane of the secretory granules fuses with the plasma membrane of the axonal bulbs, and the peptide contents are then extruded into the adjacent capillaries. The Brattleboro rat, a strain with an autosomal recessive defect that causes AVP deficiency, is afflicted by a single base deletion in exon B. This leads to a shift in the reading frame, with loss of the translational stop code. Although transcribed and translated in the hypothalamus, the translational product is neither transported nor processed in these mutant rats.

Figure 1-5 Structure of the human antidiuretic hormone, arginine vasopressin. (From Schrier RW, Miller PD. Water metabolism in diabetes insipidus and the syndrome of inappropriate antidiuretic hormone secretion. In: Kurtzman NA, Martinez Maldonado M, eds. Pathophysiology of the Kidney. Springfield, IL: Charles C Thomas; 1977.)

Figure 1-6 The arginine vasopressin (AVP) gene and its protein products. The three exons encode a 145-amino acid prohormone with an NH-2-terminal signal peptide. The prohormone is packaged into neurosecretory granules of magnocellular neurons. During axonal transport of the granules from the hypothalamus to the posterior pituitary, enzymatic cleavage of the prohormone generates the final products: AVP, neurophysin, and a COOH-terminal glycoprotein. When afferent stimulation depolarizes the AVP-containing neurons, the three products are released into capillaries of the posterior pituitary. (From Brenner BM, ed. The Kidney. 8th ed. Philadelphia: Saunders Elsevier; 2008, with permission.)

The regulation of AVP release from the posterior pituitary is dependent primarily on two mechanisms: osmotic and nonosmotic pathways (Fig. 1-7).

Osmotic Release of Vasopressin

The osmotic regulation of AVP is dependent on “osmoreceptor” cells in the anterior hypothalamus in proximity but separate from supraoptic nuclei. These cells, most likely by altering their cell volume, recognize changes in ECF osmolality. Cell volume is decreased most readily by substances that are restricted to the ECF, such as hypertonic saline or hypertonic mannitol, and thus enhance osmotic water movement from cells; these substances are very effective in stimulating AVP release. Since the effects of saline and mannitol are comparable, this supports the view that the response is due to changes in effective osmolality rather than to sodium per se. In contrast, urea moves readily into cells and therefore does not alter cell volume; hypertonic urea does not effectively stimulate AVP release. The effects of increased osmolality on vasopressin release are associated with measurable (twofold to fivefold) increases in vasopressin precursor messenger RNA (mRNA) in the hypothalamus. The osmoreceptor cells are very sensitive to changes in ECF osmolality. An increase of ECF osmolality by 1% stimulates AVP release, whereas water ingestion causing a 1% decrease in ECF osmolality suppresses AVP release (Fig. 1-8). A role for members of the transient receptor potential vallinoid family (TRPV 1 and 4) in osmoregulation has been suggested as knockouts of these proteins in mice become somewhat hypernatremic and display blunted vasopressin secretion in response to hypertonic stimuli as summarized by Cohen (12).

A close correlation between AVP and plasma osmolality has been demonstrated in subjects with various states of hydration, although there are considerable genetically determined individual variations in both the
threshold and sensitivity (Fig 1-8). In humans, the osmotic threshold for vasopressin release is between 280 and 290 mOsm/kg. The system is so efficient that plasma osmolality usually does not vary more than 1% to 2 %, despite great variations in water intake. There is also a close correlation between AVP and urinary osmolality, allowing for the maintenance of tonicity of body fluids.

Figure 1-7 Osmotic and nonosmotic stimulation of arginine vasopressin release. (From Robertson GL, Berl T. Pathophysiology of water metabolism. In: Brenner BM, ed. The Kidney. 6th ed. Philadelphia: WB Saunders; 2000:875, with permission.)

Figure 1-8 Antidiuretic hormone levels, urinary osmolality, and thirst as functions of serum osmolality. (From Narins RG, Krishna GC. Disorders of water balance. In: Stein JH, ed. Internal Medicine. Boston: Little, Brown; 1987:794, with permission.)

Nonosmotic Release of Vasopressin

Vasopressin release can occur in the absence of changes in plasma osmolality (5). Although a number of such nonosmotic stimuli exist, physical pain, emotional stress, and a decrement in blood pressure or volume are the most prominent ones. A 7% to 10% decrement in either blood pressure or blood volume causes the prompt release of vasopressin (Fig. 1-7). Because the integrity of the circulatory volume takes precedence over mechanisms that maintain tonicity, activation of these nonosmotic pathways overrides any decline in the osmotic stimulus that otherwise would suppress the hormone’s release. This process accounts for the pathogenesis of hyponatremia in various pathophysiologic states, including cirrhosis, heart failure, and several endocrine disorders.

There is considerable evidence for the existence of baroreceptor sensors in the low-pressure (venous) areas of the circulation, particularly in the atria. Atrial distention causes a decrease in plasma AVP levels, and a water dieresis; this reflex is mediated by the vagus nerve. Alternatively, arterial baroreceptors in the aorta and carotid sensors send impulses through the vagus and glossopharyngeal nerves to the nucleus tractus solitarii of the medulla. Unloading of these arterial baroreceptors decreases tonic inhibition and leads to the nonosmotic release of vasopressin. Denervation of these arterial baroreceptors has been shown to abolish the nonosmotic release of AVP.

It is possible that angiotensin II is a mediator of AVP release in these states because many of the pathophysiologic states associated with nonosmotic AVP release are characterized by enhanced plasma renin activity and therefore increased angiotensin II levels. The experimental results in this regard are however conflicting. Activation of the sympathetic nervous system seemed be involved in the nonosmotic stimulation of AVP. In this regard, the supraoptic nuclei are heavily innervated by noradrenergic neurons. Other pathways that could stimulate the nonosmotic secretion of AVP have been proposed; for example, the antidiuresis associated with nausea and pain has been ascribed to an emetic and to a cerebral pain center, respectively. A role for baroreceptor pathways however has not been convincingly excluded even in these settings. Other biogenic amines, polypeptides, and even cytokines have been implicated as modulators of AVP release in addition to catecholamines.

Cellular Action of Vasopressin

Once released from the posterior pituitary, vasopressin exerts its biologic action on water excretion by binding to V2 receptors in the basolateral membrane of the collecting duct (Fig. 1-9) (13). The receptors to which vasopressin binds have been cloned. The V1 receptor on blood vessels and elsewhere is a 394 amino acid protein with seven transmembrane domains (14). The 370 amino acid V2 receptor, which is present only in the kidney and has a similar configuration, has been cloned for both rats (15) and humans (16). Although the V1 receptor messenger is plentiful in the glomerulus, it is also detected in the collecting duct, where the V2 message is predominant.

Binding of AVP to its V2 receptor increases adenylate cyclase activity resulting in the generation of cyclic adenosine 3′,5′-monophosphate (cAMP) from adenosine triphosphate. The V2 receptor is coupled to the catalytic unit of adenylate cyclase by the stimulatory guanine nucleotide binding regulatory protein, Gs. This is a heterotrimeric protein whose α subunit binds and hydrolyzes guanosine triphosphate. The heightened cAMP formation activates protein kinase A (PKA), which in turn phosphorylates serines and threonines. The activation of PKA brings about the phosphorylation of the water channel AQP2 at serine 256 in intracellular vesicles, and thereby increases the trafficking of the water channel to the luminal membrane (17). This sequence of events results in a marked increase in the water permeability of the luminal membrane and thereby the collecting tubule. AQP2 is a member of an increasingly large family of water channels whose archetypal member, AQP1, cloned by Agre and coworkers (18) is, as mentioned previously, abundant in the proximal tubule and the proximal half of the descending limb of Henle. In contrast, AQP2 is limited to the vasopressin-sensitive principal cell of the collecting duct, and particularly to the cytoplasm and luminal membrane. Vasopressin also is involved in the long-term regulation of the expression of AQP2 (19). AQP3 and AQP4 are widely distributed, including the collecting duct principal cell, where they are localized at the basolateral membrane. In this basolateral membrane, they serve as conduits for water exit from the cell. Other AQP6 to AQP8 also are expressed in the kidney. AQP6 is present in intercalated cells, AQP7 in the S3 segment of the proximal tubules, AQP8 in proximal tubules and collecting ducts, and AQP11 in the proximal tubule (19). The physiologic roles of these water channels in the kidney are not clear. The cytoskeleton also plays an important role in the trafficking of the AQP2 water channel to the luminal membrane, a process involving both exocytic insertion associated with AVP stimulation and endocytic retrieval associated with suppression of AVP action.

Figure 1-9 Schematic representation of the cellular action of vasopressin. The binding of vasopressin to the V2 receptor in the basolateral membrane initiates a cascade of events resulting in AQP2 insertion in the luminal membrane, see text for details. (From Bichet D. Nephrogenic and central diabetes insipidus. In: Schrier RW, ed. Diseases of the Kidney and Urinary Tract. Vol 3. 7th ed. New York: Lippincott Williams & Wilkins; 2012:2553, with permission from Wolters Kluwer Health.)

Quantitation of Renal Water Excretion

The quantitation of water excretion has been facilitated by the concept that urine flow (V) is divisible into two components. One component is the urine volume needed to excrete solutes at the concentration of solutes in plasma. This isotonic component has been termed osmolar clearance (Cosm). The other component is called solute-free water clearance (image) and is the theoretic volume of solute-free water that has been added to (positive image) or reabsorbed from (negative image or image) the isotonic portion of urine (Cosm) to create with hypotonic or hypertonic urine, respectively. These terms are calculated as follows:

Further inspection of these relationships will reveal the following:

  • 1. When Uosm equals Posm (isotonic urine), V equals Cosm; therefore, image is zero.

  • 2. When Uosm is greater than Posm (hypertonic urine), Cosm is greater than V; therefore, image is negative (also denoted as image).

  • 3. When Uosm is less than Posm (hypotonic urine), Cosm is less than V, and image is positive.

This relationship is depicted further in Figure 1-10.

The excretion of hypertonic urine has the net effect of returning solute-free water to the organism and thereby dilutes body fluids. In contrast, the excretion of hypotonic urine has the net effect of ridding the organism of solute-free water and thus concentrating body fluids. Urine osmolality alone does not give the volume of water added to or removed from the organism; the calculation of image or image better allows the quantitation of water balance.

A limitation of the equation is that it fails to predict clinically important alterations in tonicity and serum sodium concentration because it factors in urea. Urea is an important component of urine osmolality, but does not establish transcellular osmotic gradients because it readily crosses cell membranes. Consequently, urea influences neither the serum sodium concentration nor the release of vasopressin, and its inclusion in urine
osmolality does not predict changes in serum sodium. This is better reflected if specifically electrolyte-free water clearance (image[e]) is measured. In this formulation, the serum osmolality is replaced by serum sodium and urine osmolality by UNa + UK.

Figure 1-10 Relationship between urine flow (V), Cosm, CH2O, and image in hypotonic, hypertonic, and isotonic urine.


If a patient’s UNa + UK < PNa, then image(e) is positive, a process that will raise the plasma concentration of sodium. Conversely, if UNa + UK > PNa, then image(e) is negative, a process that tends to lower the serum concentration of sodium.


The ingestion of a diet containing average amounts of sodium (150 mmol/day) and protein (70 g/day) has to dispose of approximately 600 mOsm of solutes per day. The daily volume of urine in which this solute is excreted depends on fluid intake. The 600 mOsm can be excreted in 6 L of urine with an osmolality of 100 mOsm/kg H2O if the daily fluid intake is generous. If water ingested is limited and renal concentrating capacity is intact, then the 600 mOsm solute load can be excreted in 500 mL of urine with an osmolality of 1,200 mOsm/kg H2O.

This flexibility in daily urine volumes for a given solute load is limited if renal concentrating ability is impaired. For example, if the maximal renal concentrating ability is reduced to 300 mOsm/kg H2O, then the 600 mOsm of solute obligates 2 L of urine per day to maintain total body solute. The 600 mOsm of daily solute requires 10 L of urine per day with a more severe concentrating defect that does not allow urine to be concentrated above 60 mOsm/kg H2O.

In terms of water conservation, the kidney’s ability to increase urine osmolality from 60 to 300 mOsm/kg H2O is quantitatively more important than its ability to increase urine osmolality from 300 to 1,200 mOsm/kg H2O. For example, with a daily solute load of 600 mOsm, a decrease in maximal urine osmolality from 1,200 to 300 mOsm/kg H2O increases obligatory urine flow from .5 to 2.0 L/day. Thus, severe polydipsia and polyuria should not be observed even in the complete absence of the renal capacity to concentrate urine above plasma. However, for the same solute load, a further decrease in maximal urinary concentration from 300 to 60 mOsm/kg H2O requires the excretion to increase from 2 to 10 L of urine/day. This degree of defect in water conservation obviously is associated with overt polyuria and polydipsia. In this setting, a severe water deficit and hypernatremia occur in the absence of an intact thirst mechanism and a large intake of water.


In quantitative terms, the normal kidney’s ability to reabsorb image is more limited than its ability to excrete image. With maximal urine osmolality of 1,200 mOsm/kg H2O and a daily urine volume of 500 mL, image can be calculated as follows:

Thus, only 1,500 mL of solute-free water is returned to body fluids during this maximal antidiuresis. In contrast, with the same daily solute load of 600 mOsm, a minimal urine osmolality of 60 mOsm/kg H2O, and a daily urine volume of 10 L, the renal capacity to excrete image is much greater than the capacity to return solute-free water (image) to the body. More specifically,

Thus, with comparable solute loads and relatively maximal and minimal urine osmolalities, the image of 1.5 L/day is substantially less than the image of 8 L/day.

Thus, prevention of a total body water deficit is largely dependent on water intake as modulated by thirst. The thirst center appears to be closely associated anatomically with the osmoreceptor in the region of the hypothalamus. Defects in thirst response may involve either organic or generalized central nervous system (CNS) lesions and can lead to severe water deficit even in the presence of a normal concentrating mechanism. Of course, the water deficit occurs more promptly if renal concentrating ability is impaired as well.

Clinical Disorders of Urinary Concentration Causing Hypernatremic States

The renal concentrating mechanism represents the first defense against water depletion and hyperosmolality. A perturbation in any component of the concentrating mechanism, shown in Figure 1-1B, culminates in an inability to maximally concentrate urine. Renal concentrating defects ensue when there is impairment in the generation of medullary hypertonicity either as a consequence of decreased delivery of solutes to the loop (diminished GFR) or inability to reabsorb NaCl in the loop of Henle (loop diuretics). Likewise, failure to render the collecting duct permeable to water because vasopressin is absent or the tubule is unresponsive to vasopressin also results in a renal concentrating defect. Thirst becomes a very effective mechanism for preventing further increases in serum sodium when renal concentration is impaired (20, 21). The plasma osmolality threshold for thirst appears to be approximately 10 mOsm/kg H2O above that of vasopressin release (Fig. 1-8). In fact, thirst is so effective that even patients with complete diabetes insipidus avoid hypernatremia by fluid intake in excess of 10 L/day. Therefore, hypernatremia supervenes only when hypotonic fluid losses occur in combination with a disturbance in water intake (22). This is most commonly seen in the aged (with an alteration in level of consciousness), the very young (with inadequate access to water), or a rare subject (with a primary disturbance in thirst).

Hypernatremia can develop with either low, normal, or, more rarely, high total body sodium, as shown in Figure 1-11 (23).


Patients who sustain losses of both sodium and water, but with a relatively greater loss of water, are classified as having hypernatremia with low total body sodium. Such patients exhibit the signs of hypovolemia such as orthostatic hypotension, tachycardia, flat neck veins, poor skin turgor, and dry mucous membranes. The causes that underlie the hypovolemic state are similar to those that cause hypovolemic hyponatremia. The effect on serum sodium is determined by the failure to ingest water (hypernatremia) or excessive free water intake (hyponatremia). Extrarenal loss of hypotonic fluid can occur either through the skin because of profuse sweating in a hot and/or humid environment or, more frequently, from the gastrointestinal tract in the form of diarrhea. Lactulose-induced diarrhea leading to hypernatremia appears to be common, although primarily recognized in children. Urine osmolality is high (usually >800 mOsm/kg H2O) and urinary sodium concentration is low (<10 mEq/L) because the renal water and sodium conserving mechanisms operate normally in these patients. Hypotonic losses also can occur by the renal route during a loop diuretic-induced hypotonic diuresis or an osmotic diuresis with either mannitol, glucose, or as is not uncommon, urea in the setting of excessive protein supplementation. Elderly patients with partial urinary tract obstruction can excrete large

volumes of hypotonic urine. The urine with such obstruction is hypotonic or isotonic, and the urinary sodium concentration is greater than 20 mEq/L. As glucose and mannitol enhance osmotic water movement from the intracellular fluid to the ECF compartment, these patients may have a normal or even low serum sodium concentration in spite of serum hypertonicity.

Figure 1-11 Diagnostic and therapeutic approach to the hypernatremic patient. (From Berl T, Kumar S. Disorders of water balance. In: Johnson RJ, Feehally J, eds. Comprehensive Clinical Nephrology. St. Louis: CV Mosby; 2000:3-9, with permission.)


Loss of water without sodium does not lead to clinically significant volume contraction unless the water losses are massive. Therefore, patients with hypernatremia secondary to water loss appear to be euvolemic with normal total body sodium. The extrarenal sources of such water losses are the skin and respiratory tract. A high environmental temperature as well as a febrile or hypermetabolic state can cause considerable water losses. Hypernatremia supervenes if such hypotonic losses are not accompanied by appropriate water intake. Urine osmolality is very high, reflecting an intact osmoreceptor-vasopressin-renal response. Urinary sodium concentration varies according to the patient’s sodium intake.

More frequently, the losses of water are of renal origin, as in diabetes insipidus. Diabetes insipidus is a polyuric disorder characterized by high rates of electrolyte-free water excretion. Hypernatremia supervenes when these losses are not appropriately replaced. Depending on whether the water losses are caused by a failure to secrete vasopressin or renal resistance to the hormone, the diabetes insipidus is designated as being central or nephrogenic, respectively.

Central Diabetes Insipidus

Failure to normally synthesize or secrete vasopressin limits maximal urinary concentration and causes varying degrees of polyuria and polydipsia, depending on the severity of the disease. The causes of central diabetes insipidus are listed in Table 1-1.

In a survey of 79 children and young adults, the disease was idiopathic in 52%, with a significant number having tumors and Langerhans cell histiocytosis. Most had magnetic resonance imaging (MRI) findings and some had thickening of the pituitary stalk that may reflect lymphocytic infiltration as part of an autoimmune process. The probability of also developing anterior pituitary hormone deficiency was 80% in the group that had tumors, compared to 50% in subjects with idiopathic diabetes insipidus (24, 25). The disease may rarely be inherited. Families with an autosomal dominant inheritance pattern have been described (26). Mutations in the coding region of the gene in all three exons have been described affecting one allele. These mutations are in the signal protein and neurophysin. Most are missense mutations, but other mutations have been described as well (26).What is peculiar about this inherited form of central diabetes insipidus is that the onset of symptoms is delayed for several months after birth and sometimes even longer. It appears that the mutant hormone forms complexes with the native hormone and the accumulation of these complexes in the endoplasmic reticulum causes progressive loss of vasopressin-producing neurons (27). There is also a rare inherited autosomal recessive form of central diabetes insipidus that occurs in association with diabetes mellitus, optic atrophy, and deafness (Wolfram syndrome). The syndrome appears to result from mutations in region PC 16 of chromosome 4, which codes for a protein expressed in various tissues (28).

Head trauma, hypophysectomy, and neoplasms, either primary or metastatic (mainly from lung and breast tumors), constitute most of the other causes. Other etiologic factors include encephalitis, sarcoidosis, eosinophilic granuloma, and histiocytosis. Finally, central diabetes insipidus has been described following development of cerebral edema in 11 postoperative hyponatremic women (29).

Clinical Features

Polyuria and polydipsia are the hallmarks of central diabetes insipidus and must be considered in the differential diagnosis of any patient who presents with such symptoms. As illustrated in Figure 1.10, polyuria can occur from a solute diuresis, in which case Cosm is increased and the urine osmolality is greater than 300 mOsm/kg. A diagnosis of central (vasopressin-deficient) diabetes insipidus should be considered when polyuria is caused by an increase in image and urine osmolality is less than 150 mOsm/kg. Urine flow can range between 3 and 15 L/day, depending on the severity of the disease. The disorder frequently has an abrupt onset and occurs with equal frequency in both sexes. Although the time of onset is extremely variable, it is rare in infancy and is most frequent in the 10- to 20-year age group. Patients with central diabetes insipidus often have a predilection for cold water. Nocturia frequently is marked because there is little diurnal variation in the polyuria. Bladder
capacity may be increased in untreated patients, however; consequently, nocturia may not be a prominent symptom. Nevertheless, nocturia is frequent generally, and sleep deprivation commonly leads to fatigue and irritability. Patients with central diabetes insipidus do not develop hypernatremia if the thirst mechanism is intact and water is available; thus, they have no symptoms except for the inconvenience associated with marked polyuria and polydipsia. However, severe and even life-threatening hypernatremia can supervene with concomitant hypodipsia, no access to water, or an illness that precludes adequate water intake.

Table 1-1 Causes of Central Diabetes Insipidus


Autosomal dominant

Autosomal recessive (Wolfram syndrome)


Head trauma, skull fracture, and orbital trauma


Suprasellar and intrasellar tumors

Primary (suprasellar cyst, craniopharyngioma, pinealoma, meningioma, and glioma)

Metastatic (breast or lung cancer, leukemia, and lymphomas)



Wegener granulomatosis




Eosinophilic granuloma

Hand-Schüller-Christian disease




Guillain-Barré syndrome


Cerebral aneurysm

Cerebral thrombosis or hemorrhage

Sickle cell disease

Postpartum necrosis (Sheehan syndrome)

Pregnancy (transient)

From Levi M, Berl T. Water metabolism. In: Gonick HC, ed. Current Nephrology. Vol 5. Chicago: Year Book Medical Publishers; 1982:23, with permission.

Treatment of Central Diabetes Insipidus

Patients with central diabetes insipidus do not develop hypernatremia if the thirst mechanism is intact and water is available; thus, they have no symptoms except for the inconvenience associated with marked polyuria and polydipsia. Nonetheless, studies in acute care settings have clearly shown that patients who present with (34) or acquire (35) hypernatremia have a higher risk for mortality.

Both AVP replacement and pharmacologic agents are available for the treatment of central diabetes insipidus (Table 1-2). In acute settings, such as after hypophysectomy, the aqueous vasopressin (Pitressin) preparation
is preferable. Its short duration of action allows for more careful monitoring and decreases the likelihood of complications such as water intoxication.

Table 1-2 Therapeutic Regimens for the Treatment of Diabetes Insipidus



Complete central diabetes insipidus


10-20 µg intranasally q12-24 h


100-800 µg/d

Partial central diabetes insipidus


As above

Aqueous vasopressin

5-10 units SQ q4-6 h


250-500 mg/d


500 mg t.i.d.-q.i.d.


400-600 mg/d


Thiazide diuretics

Amiloride (for lithium-related NDI)

5 mg q.i.d.

Gestational diabetes insipidus


As above

DDAVP, desmopressin acetate; NDI, nephrogenic diabetes insipidus.

Reprinted from Thurman JB, Berl T. Therapy in nephrology and hypertension. In: Wilcox JN, ed. Therapy in Nephrology and Hypertension. 3rd ed. Philadelphia: Saunders; 2008:337-352, with permission from Elsevier.

A modification of the natural vasopressin molecule to form desmopressin acetate (DDAVP) has resulted in a compound with prolonged antidiuretic activity (6-24 hours) and virtual elimination of V1 vasopressor receptor activity (antidiuretic to pressor ratio of approximately 2,000:1) as compared with the natural hormone AVP (duration of action of 2-4 hours and antidiuretic to pressor ratio of approximately 1:1). Substitution of D-arginine for L-arginine at position 8 resulted in a peptide DAVP with diminished vasopressor activity, and deamination of the homocysteine at position 1 gave rise to a second peptide, with enhanced antidiuretic pressor activity and prolonged duration of action. Desmopressin acetate is administered intranasally in a dosage ranging from 10 to 20 µg every 8 to 12 hours. The drug has eliminated the need for previously employed long-acting vasopressin in oil. There are considerable individual variations in the required dosage, but most patients require twice-daily administration for good control of polyuria. Desmopressin acetate also can be administered intravenously or subcutaneously during periods of respiratory illness or surgery.in doses between 1 and 4 g. It is active orally in large doses also (between 50 and 800 µg) (34).

Large doses of DDAVP may cause transient headaches, nausea, and a slight increase in blood pressure; these symptoms disappear if the dosage is reduced. Nasal congestion, mild abdominal cramps, and vulval pain have occurred rarely. These patients need careful monitoring of water intake and serum sodium to avoid development of hyponatremia. In fact, there are increasing reports of cases of hyponatremia in patients on these agents, particularly when used for other indications such as von Willebrand disease (35) and enuresis (36).

Intranasal DDAVP currently is the treatment of choice for partial or complete central diabetes insipidus. However, alternatives to hormone replacement may be helpful at times. With dilute urine of fixed low osmolality, the urine volume is determined by the solute load requiring excretion. A reduction in salt and protein in the diet therefore will reduce the major urinary solutes and thus the volume of urine necessary to accommodate their excretion. Moreover, a number of pharmacologic agents with antidiuretic properties are used; the hypoglycemic agent chlorpropamide (Diabinese) is the most commonly employed. Its antidiuretic effects are manifested only if some vasopressin is present; therefore, it is useful only in partial diabetes insipidus. A trial of 250 mg every day or twice a day may be offered to patients with partial central diabetes insipidus and at least 7 days allowed for an effect to occur. The anticonvulsant carbamazepine (Tegretol) has caused antidiuresis in subjects with central diabetes insipidus. A combination of chlorpropamide and carbamazepine has been found to provide an effect that could be synergistic. Clofibrate also has been used to treat partial central diabetes insipidus. At present, however, none of these approaches can be recommended over intranasal DDAVP.

Congenital Nephrogenic Diabetes Insipidus

Congenital nephrogenic diabetes insipidus is a rare hereditary disorder in which the renal tubule is insensitive to vasopressin (37). The disease has been described in various patterns, including the X-linked form, and autosomal recessive and even an autosomal dominant form. The most common variety is the X-linked whose complete form manifests itself in males with females expressing variable degrees of polyuria and polydipsia. In 85% of patients, the disease is a consequence of mutations on the V2 receptor, resulting in a loss of function (38). More than 180 mutations of the V2 receptor in chromosome region Xq28 have been identified (39). Half are missense mutations but other types of mutations occur as well. A significant number of the mutant receptors have defective intracellular trafficking (37). The autosomal recessive form of congenital nephrogenic diabetes insipidus is to a mutation in the AQP2 water channel and accounts for approximately 15% of disease (40). At least 30 disease-causing mutations have been identified, most of them of the missense type. As was the case with the V2 receptor mutant, misrouting of the AQP2 mutant protein has been described in this setting (41). Mutations at the carboxy terminal of AQP2 can cause a rare autosomal dominant form of congenital nephrogenic diabetes insipidus (42). A modest concentrating defect is present also in humans deficient in the AQP1 water channel (43). Finally, knockout mice lacking AQP3 and/or AQP4 also fail to maximally concentrate their urine (44).

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Nov 17, 2018 | Posted by in NEPHROLOGY | Comments Off on Disorders of Water Homeostasis

Full access? Get Clinical Tree

Get Clinical Tree app for offline access