Renal handling of water

This chapter seems relevant to the aims of Section H1(vi) from the 2017 CICM Primary Syllabus, which expects the exam candidate to "describe the role of the kidneys in the maintenance of fluid... balance". In the case of the homeostatic mechanisms which maintain water balance in the body, the role of the kidney can be described as wholly subservient to the hypothalamic-pituitary-adrenal axis. The kidney is an effector instrument, and does what it is told by its neurohormonal masters. It only has the most rudimentary mechanisms of regulating fluid balance by itself, those being the crude levers of renal blood flow autoregulation and tubuloglomerular feedback.

Several water-handling questions have appeared through the historical CICM exam papers:

• Question 9 from the second paper of 2018
• Question 4 from the first paper of 2015
• Question 8 from the first paper of 2008

In summary:

Or, in the form of a huge diagram with unreadably small writing,

Weirdly, for a series of mechanisms which seem so central to human physiology, there are surprisingly few definitive articles in print. James Schafer's 2004 retrospective is probably the easiest to read, but does not cover all the detail. Lassiter et al (1961) covers a lot of detail, and has an enviably logical structure, but is so ancient that the authors were only able to present experimental results and speculations. 

Reabsorption of water in the proximal tubule

Water is reabsorbed passively along the length of the nephron. It is filtered in the glomerulus at a rate of perhaps 200 ml/min, and enters the proximal tubule along with everything else. Here, the powerful gradients generated by the basolateral Na+/K+ATPase pumps create a gradient to drive the reabsorption of sodium, and the reabsorption of sodium is then used to power all the cotransporters and exchange pumps to reclaim all the other solutes. 

The net effect of all this solute reabsorption is a decrease in the osmolality of the tubular fluid, which becomes hypotonic in comparison to the extracellular fluid beyond the tubule. At the same time, the fluid in the basolateral intercellular space becomes hyperosomolar, as the solutes being secreted from the basolateral membrane of the proximal convoluted tubule cells end up concentrated in the interstitial fluid. Bishop (1978) reported that the measured osmolality of this renal cortical fluid is about 150 mOsm/kg higher than that of plasma. Well, they centrifuged that fluid out of some macerated rat kidneys, but it's probably close enough for a ballpark figure.

This produces an osmotic gradient. Now, the proximal tubule is insanely water-permeable - much more so than the collecting duct or the thin descending limb.  In fact, because of its huge surface area (microvilli) and abundant transcellular (aquaporins) and paracellular (tight junctions) transport options, the permeability of the proximal tubule is several orders of magnitude greater than that of the other tubules. Berry, 1983, gave a figure of something like 0.2 cm/s for the proximal tubule, whereas Morgan & Berliner in 1968 reported the water permeability coefficient of the thin limb as 0.0012 cm/s.  That means there is basically no barrier.  Water is quite free to follow that osmotic gradient out of the proximal tubule until it has equilibrated with the extracellular fluid. No matter how much solute is reabsorbed, the tubular fluid will remain isoosmolar. 

Uptake of renal cortical water by peritubular capillaries

The volume of the water reclaimed by the proximal tubule (75-80% of the total filtered volume) would end up being 160ml/min, so something must clearly be removing this extra water from the renal cortex. It ends up being sucked up into the cortical peritibular capillaries. This water movement is driven by several factors, which are beautifully described by Aukland et al (1994). Unsurprisingly, the best way to describe them is using some sort of simplified form of the Starling equation:

Jv = L_p S [ (P_c - P_i) - (Π_c - Π_i) ]; 

where

•  P_i is the interstitial hydrostatic pressure in the renal cortex. Because this osmotic gradient creates a constant water flow from the proximal tubule lumen, and because the renal capsule is not particularly stretchy or compliant, this water movement has the effect of increasing the volume of the cortical interstitial fluid, and consequently its pressure. It's not much - Aukland et al (1994) reported a value of 2-4 mmHg - but it is enough to contribute to the Starling equation. Given that the kidney is not a particularly distensible organ, one should expect this pressure to change quite dramatically from even minor fluctuations  in interstitial fluid volume.
• P_c is the pressure inside the peritubular capillaries, which is really rather low - 8 mmHg according to Aukland et al.
• Π_i is the interstitial oncotic pressure, which under conventional circumstances is close to zero (as this fluid is usually rather protein-poor)
• Π_c is the capillary oncotic pressure, which is about 34 mmHg - higher than normal  because of the haemoconcentrating effect of glomerular ultrafiltration
• L_p S is the permeability coefficient and surface of the capillary surface, which for peritubular capillaries are highly favourable to water movement (i.e. vast in area and highly fenestrated)

Or, in picture form:

Thus, the pericapillary fluid which surrounds the peritubular capillaries is under considerable oncotic  pressure (and trivial hydrostatic pressure) both of which are constantly sucking it into the lumen of the capillary. The forces which oppose this (minimal oncotic pressure of the interstitial fluid and the paltry capillary hydrostatic pressure) are insufficient to prevent water movement. The net effect is a highly efficient and self-regulating mechanism of fluid removal which requires little added energy to move vast amounts of water, beyond what has already been expended by the basolateral Na⁺/K⁺ ATPase.

Reabsorption of water in the loop of Henle

On the way out of the proximal tubule, the isoosmolar tubular fluid descends along the thin limb of the loop of Henle into the briny depths of the medulla. The thin descending limb is even more permeable to water than a vasopressin-treated cortical collecting duct (Morgan & Berliner, 1968), though not as permeable as the proximal tubule. The increasing osmotic gradient pulls water out of the descending thin limb and into the medullary intersititum, where it is immediately dragged away by the ascending vasa recta in the course of countercurrent exchange. The ascending vasa recta are of course carrying blood which has just spent some time absorbing all the solutes in the renal medulla, and so at every level of the medulla the blood inside these vessels is hyperosmolar in comparison to the medulary interstitium. Therefore, at any given level and interstitial osmolality value, the medullary interstitial water will be osmotically attracted into the ascending vasa recta.

In this fashion, the renal medulla is able to reclaim another 10-15% of the glomerular filtrate volume. The exact percentage is not something the trainee should fixate on, as it is meaningless and clearly different depending on what kind of nephron you are looking at. 

With the tubular contents losing water to the hyperosmolar medulla, at the hairpin bend of the loop we end up with a fluid which is markedly hyperosmolar (1200-1400 mOsm/kg), with a sodium concentration of roughly 250 mmol/L and potassium near 50 mmol/L. From here, as this fluid ascends via the thin and thick ascending limbs, a lot of these electrolytes will be removed by active transport, leaving being fluid which will be markedly hypotonic, with an osmolality as low as 90 mOsm/Kg.

Water in the distal convoluted tubule

The distal convoluted tubule is said to be water-impermeable. This short segment of the nephron is not interested in helping you with your water balance problems until you threaten it with thiazide diuretics. Over the length of the distal convoluted tubule, Clapp & Robinson (1966) demonstrated a roughly unchanged osmolality of DCT fluid, which remained hypotonic in relation to plasma (somewhere around 60-120 mOsm/kg). Only towards its most distal end, where it blends with the connecting tubule, does this segment begin to express aquaporins.

Reabsorption of water in the collecting duct

Aquaporin expression is a defining feature of collecting duct prinicipal cells. These are transmembrane proteins which insert into the apical and basal membranes of the collecting duct. As the duct dives deeper into the hyperosmolar renal medulla, the dilute output of the distal convoluted tubule can be concentrated by the osmotic gradient.

The expression of these aquaporins represents a very important regulatory lever, pulled by the hypothalamic-pituitary axis through the secretion of vasopressin. Loss of vasopressin binding at the V2 receptors in this segment gives rise to a decreased expression of apical aquaporins, and nothing gets reabsorbed. The end product of the collecting duct ends up being urine at basically the same concentration as it was when it left the distal convoluted tubule, i.e. something like 50-60 mOsm/kg. But flush the duct with vasopressin, and its water permeability will increase so much that complete equilibration between the duct fluid and the inner medulla can take place, which produces urine with an osmolality of 1200-1400 mOSm/kg. 

Regulatory influences on the renal reabsorption of water

Every physiological system seems to have its fingers in this pie, and one day the CICM examiners may, in a fit of violent rage, ask the trainees to "describe the neurohormonal regulation of renal water handling", or something equally terrible. For that dark day, we should prepare with a solid classification system:

• Regulation of renal blood flow and glomerular filtration
  • Autoregulation
    • Myogenic response (50% of the total autoregulatory response)
    • Tubuloglomerular feedback (35%)
  • Neurohormal 
    • Sympathetic regulation 
    • Angiotensin II
    • Endothelin
    • Prostaglandin (PGE₂)
  • Substrate-dependent
    • Protein ingestion and amino acid infusion
    • Glucose (hyperglycemia)
• Regulation of tubular water permeability
  • Vasopressin:
    • V2 receptors expressed on the surface of DCT and collecting duct cells
    • Aquaporin expression is enhanced by V2 activation, facilitation reabsorption of tubular fluid
• Regulation of renal solute handling:
  • Aldosterone:
    • Increased absorption of sodium in distal tubule and collecting duct by activation of the ENaC channel
    • Therefore increased gradient for water reabsorption, provided aquaporin expression permits it (Scott et al, 2017)
  • Natriuretic peptides:
    • Decrease Na/K ATPase activity in the tubules
    • Thus, decrease the gradient for sodium reabsorption
    • This will decrease the movement of water through segments of the nephron which depend on sodium gradients.