Renal handling of urea and glucose

This chapter is relevant to the aims of Section H1(vii) from the 2017 CICM Primary Syllabus, which expects the exam candidate to "Describe the role of the kidney in the handling of glucose, nitrogenous products and drugs".  The drugs a whole separate topic and would need to be treated separately, but the handling of urea and glucose can probably be lumped together into a single chapter. Urea will obviously be the main course, as it is an important facet of renal function and a subject of at least one past paper SAQ (Question 8(p.2) from the first paper of 2010). Well, not really the subject; the question asked "explain the role of urea in the body", which sounds mainly like an invitation to describe the urea cycle. But somehow the expected answer did include some renal elements, and to be fair the kidneys and urea are fairly well alloyed in the mind of the critical care practitioner, so here we are.

In short:

For urea, you can't do much better than Klein et al (2011), but it is unfortunately paywalled. Without bypassing that barrier by some mechanism, one would have to resort to free articles such as Bankir et al (1996), Bankir (2012),  Weiner et al (2015) or Yang (2012) (which seems to be an entire full-text textbook somehow published online, presumably in a totally legal fashion). For glucose, one could refer to Mather & Pollock (2011), but realistically their abstract would probably be enough. It almost seems like it would be more important to read about SGLT2 inhibitors instead.

Physical and chemical properties of urea

It is not the intention of the author to digress extensively on the properties or metabolic roles of urea and glucose here, but they do deserve some mention here because the physics and chemistry of these molecules is fairly central to their behaviour in the renal tubule. For example, take urea. It is a small molecule, with a molar mass of 60.6g and the formula CH₄N₂O - basically two amino radical groups strapped to a carbonyl core. Under normal physiological conditions it is polar and is therefore only able to cross lipid bilayers with some difficulty, unless there are specific transport proteins fussing over it. Stepping carefully around the gaping rabbithole of urea cycle physiology, the author will exert considerable self-control by mentioning only that urea is the end product of that cycle, and plays an essential role in the removal of fixed nitrogen (or "nitrogenous wastes") in all metazoans. Specifically, in humans it is the main mechanism for ammonia disposal (renal tubular synthesis being a minor player in terms of raw ammonia volume). 

Outside of horrific eating-contest-level protein excess, the average 70kg reader on a normal Western diet would produce about 420 mmol/day (25.5g) of urea, pretty much all of which is excreted in the urine (Rudman et al, 1973). Compared to the paltry 30-50 mmol/day of ammonium ion excretion by the kidneys (in the course of normal urinary acid-bas balancing), this mechanism of nitrogen removal is clearly dominant. Both mechanisms are also obviously dependent on the function of the kidneys, and under conditions of renal dysfunction, urea can be expected to accumulate and its levels will rise by about 10mmol/L/day (assuming it distributes evenly into about 40L of body water).

"Uraemia", the state of toxicity associated with renal failure, is generally associated will all sorts of unpleasant clinical features, but for these features urea is usually not to blame, and by itself it is a rather benign molecule. At high concentrations, it does have some pharmacological effects, most of which can be characterised as inhibition of enzymes through the carbomylation of proteins. In the laboratory, this manifests as inhibition of Na-K-2Cl co-transporter in the red blood cells, or nitric oxide synthase inhibition. However, clinically, these effects are usually not especially noticeable. Cruel human experiments from forgotten decades demonstrate the relative safety of urea in human physiology. For instance, Johnson et al (1972) convinced a group of chronic renal failure patients to undergo urea loading by the addition of pure urea to their dialysate. "It was not until a value in excess of 300 mg. per 100ml. of urea was obtained that vomiting became a constant feature", the bastards calmly observed. They only went up as high as 400mg/100ml plasma urea concentration, which is about 143 mmol/L, and would have gone higher if the patients had not started complaining of headaches and somnolence. These nightmarish data suggest that the urea molecule itself, like many others, only becomes toxic or dangerous when it accumulates in truly preposterous concentrations, not seen outside of mad scientist laboratories.

So, if urea really isn't that bad, why are the kidneys so hell-bent on removing it from the body? Well, it is really ammonia they are trying to remove. Ammonia is much more chemically active and toxic, and it is produced in vast amounts in the course of normal protein metabolism, which makes its removal very important. Aquatic animals have the luxury of allowing the ammonia to diffuse away into the water, by which they are generally surrounded. Terrestrial animals only have their urine to work with. Sure, you might say, couldn't they just excrete the ammonia directly into the urine? Yes, reader, they could do that, but your own daily urine production is about 2L, which means you'd be concentrating the ammonia up to 210 mmol/L (3.57g/L), or 0.36% by weight - about the same the ammonia concentration as a normal household disinfectant. This could theoretically make for a truly spotless urinary tract, but as it is not made of porcelain the more likely outcome would be some kind of upsetting urological nightmare. Of course, instead of excreting a hideously corrosive concentrated ammonia solution, you could excrete a dilute solution by accompanying the ammonia with many litres of urine, but this would be a massive waste of water. Therefore, the only sensible option for terrestrial metazoans was to package the ammonia into an inoffensive form, so they can safely concentrate it and excrete it along with some minimum of water. Ergo, the urea cycle is necessary,  urea is necessary, and urinary elimination of urea is necessary. 

Fate of urea in the proximal tubule

Urea is filtered freely in the glomerulus. The concentration of urea entering the proximal tubule is therefore exactly the same as the plasma concentration of urea. It then becomes concentrated in the first few millimetres of the proximal tubule, mainly because of the fact that water is reabsorbed there. The concentration of urea in the late proximal strauight tubule ends up being about 50% higher than plasma, according to some micropuncture studies by Ullrich et al (1963). This is slightly less than what one might expect from straight concentration. Consider: the proximal convoluted tubule removes 65% of the total glomerular filtrate, which means the concentration of any residual solute should more than double. The reason it does not double is because some urea is reabsorbed along with water in the proximal tubule by solute drag, which is apparently quite unavoidable and not affected by any major regulatory mechanisms. In this uncontrolled fashion, about 40% of filtered urea is reabsorbed.

The urine then enters pars recta, where there is some active secretion of urea. How much, and how relevant, is a matter of some scholarly debate. Klein et al (2011) dismiss this aspect as meaningless in the grand scheme of urea elimination machinery, as the active transport here occurs "at low levels that may not be physiologically important". Other authors (eg. Bankir & Yang, 2012) rail extensively on the importance of urea transport proteins, pointing to knockout mice who had considerable difficulty maintaining an efficient medullary countercurrent exchange without UT-B transporters. It makes a bit of sense, as any increase in urea concentration at this stage (however minor) will be amplified markedly by the concentration processes upstream. For the exam-going CICM trainee the wisest option would be to stay out of these arguments and write something noncommital, like "UT-B transporters in pars recta eliminate urea by active transport". 

Progress of urea in the loop of Henle

As urea enters the thin descending limb, water drains out of the tubular lumen, progressively concentrating the urea and other electrolytes. Additionally, some urea is added to the fluid by some type of active or facilitated transport mechanisms.  Pennel et al (1974) determined that the concentration of urea in the hairpin tip of the loop of Henle was actually much higher than can be explained by purely passive mechanisms. For example, sodium ends up being concentrated passively from 140 mmol/L up to about 350 mmol/L (2.5 times the plasma concentration), whereas urea increases by up to 20-30 times,  up to around 150-200 mmol/L (Gottschalk et al, 1963, got hamster tubule results of up to 300 mmol/L). The reason for this is probably some active transport of urea by UT-A2 in the thin descending limb. The most likely source of this urea is the ascending vasa recta, which is bringing concentrated urea-rich blood back out of the inner medulla. This is a part of the mechanism of urea recirculation, which is discussed elsewhere. Without digressing much on the process of countercurrent exchange, it will suffice to say that urea is trapped in the inner medulla by a range of such catch-and-return mechanisms. By borrowing hamster numbers from Gottschalk et al (1963), rat data from Jamison et al (1967) and computer model data from Stewart (1975), some numbers (urea concentrations in mmol/L) could be added to the diagram below, with the intention of demonstrating urea concentration gradients along the nephron:

 In wordy point form:

• In the thin descending limb,
  • Urea permeability is low, and water permeability is high
  • Water is reabsorbed, and urea is massively concentrated
  • Additionally, urea originating from the ascending vasa recta is transported from the medullary intersittium into the lumen by the activity of UT-A2 urea transporters 
  • Thus, the tubular concentration of urea here exceeds that of the inner medullary interstitium 
• In the thin ascending limb:  
  • Urea permeability is much higher than in the thin descending limb, and water permeability is lower
  • U​​​​​​rea diffuses out into the medullary interstitium here
    • From there, it is captured by the vasa recta
    • Then, this urea ends up endlessly recirculating between the ascending and descending vasa recta in the countercurrent exchange mechanism
  • Urea also diffuses into the thin ascending limb, mainly by being recycled out of the collecting duct (where the urea concentration is even higher).
  • The net result of these movements is an increase in the concentration of urea by about 20-25% between the bend of the loop and the end of the thin ascending limb (Marsh et al, 1970)
  • The urea concentration in the thin ascending limb equilibrates with the medullary interstitium and by the beginning of the thick ascending limb
• In the thick ascending limb: 
  • Urea permeability is poor in the early thick ascending limb, but increases slightly towards the end. When Knepper et al (1983) perfused the cortical thick limbs of rabbits with different urea concentrations, they found that creating a urea gradient across the tubular epithelium resulted in the movement of urea out of the tubule. 
  • Some of the urea reabsorbed here ends up being transported back into the descending thin limb, and ends up trapped in a countercurrent exchange mechanism

What exactly happens to urea towards the end of the thick ascending limb is somewhat unclear from the published sources. For example, Weiner et al (2015) wrote that the urea concentration should increase:

"there is an overall increase in urea concentration in the lumen from the beginning of the thick ascending limb to the distal convoluted tubule."

On the other hand, official nephrology textbooks take a completely different view. For example, Vander's Renal Physiology (9th ed, p.75) state authoritatively that  

"beginning with the thick ascending limb and continuing all the way to the medullary collecting ducts, ... the luminal membrane permeability ... is essentially zero",

That would suggest that the urea concentration there should remain essentially unchanged (water permeability of the thick ascending limb being so poor). Completely contradicting this statement, Makanjuola & Lapsley (2014) state equally authoritatively that

"the thick ascending limb of the loop of Henle is permeable to urea (and sodium) but not to water". 

Lastly, the end-tubular osmolality here is supposed to be quite low (as low as 90 mOsm/kg according to Clapp & Robinson, 1966). This means it could not possibly contain the exact same urea concentration as the end of the thin ascending limb (where by all accounts urea concentration can be as high as 100mmol/L). There is also sodium to consider (30-60mmol/L) and all the anions required to balance it, to say nothing of things like glucose and creatinine. 

So, what are we supposed to do with all this? The ICU trainee, despaired of finding any grains of truth in this contradictory mess, must be reminded of two things. One is that, if the people writing exams questions have insight into this topic, they would immediately see this quagmire and carefully lay a course around it. It is in the primary examiners' own best interest to never ask any questions which do not have an established correct answer. The other point is that data regarding tubular concentration of anything seem to mainly be derived from whatever random rodent happens to be at hand in the laboratory. After reflecting that maximally concentrated rat urine can be twice as concentrated as maximally concentrated human urine, and other rodents even more so (chinchilla, 7600 mOsm/kg), one begins to regard these measured values and authoritative statements with the angry squint of scepticism. Fortunately, in the currently preferred cram and dump model of assessment design,  the CICM trainee does not need to think too hard about these problems. It will suffice to remember that Vander's said the thick ascending limb is urea-impermeable.

Elimination and reabsorption of urea in the distal nephron

Most sources seem to agree that there is little urea absorption along the distal convoluted tubule and in the early collecting duct. This produces a massive concentration of urea towards the end of the collecting duct, as the water is removed from the urine in the aquaporin-rich later parts of these tubules. 

Then, in the distal collecting duct (deep in the inner medulla) the terminal collecting duct becomes urea-permeable. At this stage, the urea concentration gradient forces some of the urea into the renal medullary interstitium, increasing the interstitial osmolality and helping concentrate the urine in the loop of Henle.

This means that the total urinary urea concentration (and therefore the gradient driving terminal duct reabsoprtion of urea) is largely determined by the actions of vasopressin. In the presence of a vigorous vasopressin response, the urine delivered to this area of the nephron will be highly concentrated, which means the urea concentration will be very high. Therefore there will be more diffusion of urea into the medulla out of the cortical duct, and therefore the fluid in the thin limbs of the loop of Henle will become more concentrated along its journey through the medulla. In short, this mechanism increases the efficiency of water removal from the tubular fluid, and contributed to the defence of volume and osmolality.

With these steps, even humans can achieve a urinary concentration of urea which is almost 100 times the plasma concentration. When Miles et al gave dehydrated human volunteers "pitressin" in 1954, they recorded some maximum values in excess of 1400 mOsm/kg, containing something like 800 mmol/L of urea. That probably represents some sort of maximum (a normal urinary urea value is closer to 285 mmol/L, according to Yang & Bankir, 2005), though Schmidt‐Nielsen et al (1948) report 1000 mmol/L as the top limit. This is nothing compared to rats (routinely around 700 mmol/L) and mice (1,800 mmol/L), but the absolute champions of urea concentration are xercoles such as the kangaroo rat who are able to reach urinary urea concentrations in excess of 3,600 mmol/L.

Renal handling of glucose

If this renal-flavored chapter is not the right place to talk about the metabolic relevance of urea, then to discuss glucose is even further out of the question. Let us limit ourselves to soberly noting that glucose is a hexose monosaccharide sugar which is naturally most abundant as its D-stereoisomer, giving rise to the term "dextrose".  It has a molar mass of 180 Daltons (three times higher than urea) and its pKa is about 12, which means it does not dissociate at physiological pH. It is a polar molecule, not protein bound, and dissolves in lipid with only the greatest reluctance. All of this means that in the nephron this little molecule should be expected to filter effortlessly through the glomerulus and then require specific transport mechanisms to reclaim it from the urine.

And it is essential to reclaim it. Unlike urea, the role of which is purely excremento-janitorial, glucose has the noble task of rewarding the human organism with energy. At a BSL of 5mmol/L and a daily glomerular filtration rate of 200L, one is expecting to lose 1 mole of glucose per day, equating to 180g (697 calories). Obviously this does not happen. The reason for your puzzling failure to excrete an entire cake worth of sugar every day is the tireless work of proximal tubular transporters which keep reabsorbing glucose at the same rate as it is filtered.

Glucose is reabsorbed by two main transport proteins, SGLT1 and SGLT2. Both leverage the sodium gradient to co-transport glucose out of the lumen and into the proximal tubule cells. From there, it diffuses passively though GLUT channels and is reclaimed by the peritubular capillaries. SGLT1 is present lower in the tubule then SGLT2 and has about ten times greater affinity for glucose (Mather & Pollock, 2011). However, there are many more SGLT2 transporter proteins and this species is responsible for hoovering up 90% of the available filtered glucose, with the distal SGLT1 taking care of the remaining 10%. 

This reabsorption process is saturable. It appears that under normal circumstances (i.e. where the glomerular filtration rate is normal) the threshold is 11 mmol/L, beyond which glycosuria develops.  As glomerular filtration rate decreases, so the reabsorptive capacity of the remaining functional nephrons seems to suffer, and glycosuria develops with a progressively lower BSL, or so it is thought. Weirdly, increasing the glomerular filtration rate also increases the threshold - i.e. with a supranormal GFR,  Kwong & Bennet (1974) got dog tubules to reabsorb much more glucose. The precise mechanism which explains this renal glucose reabsorption threshold is unknown, and moreover people are not entirely convinced there even is such a threshold, as glucose is still found in the urine even at normal BSL values (Wolf et al, 2009).

To stretch the normal understanding of the word "handling", it is probably also worth noting that the kidney can be a net producer of newly synthesised glucose. The proximal tubule is the only other tissue in the body (other than hepatocytes) that is capable of this. Remarkably, it has some considerable capacity to create glucose from glutamine and lactate, especially under conditions of acidosis, and can account for up to 40% of total systemic gluconeogenesis under conditions of fasting (Legouis, 2020). After prolonged starvation, this fraction can increase up to 50%. Lactate metabolism seems to account for about half of this new glucose, and the rest comes from alanine glutamine and fructose.