Distribution and regulation of potassium in the body fluids

This chapter is relevant to Section I3(ii) of the 2017 CICM Primary Syllabus, which expects the exam candidates to "describe the function, distribution, regulation and physiological importance of sodium, chloride potassium, magnesium, calcium and phosphate ions". Of all the electrolytes, potassium appears to be the favourite of the CICM examiners, perhaps because their interest is proportional to lethality. Potassium-pertinent questions pepper the past papers:

• Question 20 from the second paper of 2013 (management of hyperkalemia)
• Question 12 from the first paper of 2013 (role, distribution and regulation)
• Question 4 from the first paper of 2012 (distribution)
• Question 3(p.2) from the second paper of 2009 (management of hyperkalemia)
• Question 8(p.2) from the second paper of 2007 (consequences of hyperkalemia)

Taking no further liberties with the patience of the reader, the following summary should cover most of the expected points:

Palmer (2015) is a good overview of potassium regulation, as is Gumz et al (2015). Handling of potassium elimination is basically 95% renal, which means this material is all just a duplication of the chapter on the renal handling of potassium, and you can find some peer-reviewed reference over there. For transcellular potassium flux, nothing beats the laconic directness of Na (2005). Remarkably, a good chapter on this topic can also be found in one of the official CICM textbooks (Guyton & Hall, 13th ed., p. 389).

Distribution of potassium in the body fluids

According to Rastegar (1990):

• Total body potassium is 40-55 mmol/kg
• Of this, 90% is in the intracellular fluid. This is the only exchangeable potassium.
• Extracellular fluid contains 2% of the total body potassium, and bone contains 8%
• Thus, a 70kg person has about 3675 mmol potassium in their cells, and about 50-75 mmol of potassium in their extracellular fluid.

Potassium equilibrates freely and rapidly across the extracellular fluid. There isn’t enough of it to matter seriously in any of the Gibbs-Donnan effects, or to contribute significantly to the various osmolar forces. But yes, in order to remain completely honest, one would have to say that there is slightly less potassium in the extracellular fluid, because the Gibbs Donnan Factor for monovalent cations is approximately 0.95.

Intracellular potassium concentration is usually very high. Without digressing overmuch on the subject of intracellular fluid content, it will suffice to say that potassium vastly outnumbers other electrolytes in the cell, and its concentration there ranges from 120 to 150 mmol/L.  This concentration inside the cell is kept artificially high by the action of Na⁺/K⁺ ATPase, which exchanges 3 sodium atoms for every 2 potassium.  The skeletal muscle contains most of one's potassium stores, because it contributes the largest intracellular volume to the overall count of cells (this becomes relevant later, when we discuss forcing potassium into and out of cells).  

Physiological role for potassium

"Has many important roles" the examiners said in their answer to Question 12 from the first paper of 2013, because presumably they did not want to give away that, in truth, we do not know what the role is for most of these electrolytes, other than that they are clearly present in all living things and therefore must be somehow essential. This question is hard, like all the other questions where you take a clearly essential elemental part of life and question its purpose. What is the physiological role of hydrogen? Of the neutron? Weak atomic force? Most published authors don't put this question to any sort of scrutiny, probably because most people would agree it does not need to be asked. The only time it comes up is in exams, and whenever somebody decides to write some kind of reference textbook entry on potassium. Thus, for example, Deshpande et al (2013) from The Encyclopedia of Metalloproteins felt it necessary to produce "a comprehensive reference of the role of potassium ions in biological system with examples". Unfortunately, under the "mammal" section, the discussion of "physiological roles" mainly revolves around the distribution and regulation of potassium. Getting regulated is not a role. 

Trying to make sense of this from the viewpoint of "what would the CICM examiners want here", the following list of "roles" could be concocted:

• Maintenance of intracellular fluid tonicity, thus
  • Regulation of cell volume
• Maintenance of resting membrane potential; thus:
  • Responsible for the excitability of excitable tissues
  • Critical for the transmission and generation of action potentials
  • Thus, essential for neuromuscular, cardiac and neuronal cell function
• Structural function
  • Incorporated into bone 
  • Integral part of functional ribosomes
  • Essential cofactor for the stability of DNA and RNA
• Intracellular messenger functions
  • Activation of intracellular enzymes
  • Regulation of internal cellular pH
• Extracellular messenger
  • Eg. mediator of nociception, inflammation, vasodilation

Probably a more interesting question would be, how did potassium end up the dominant intracellular cation, and sodium the dominant extracellular cation, and why is this relationship conserved throughout the entire range of living things? That's a thought probably outside of the normal psychological range of the CICM First Part Exam candidate, whose raw animal survival instincts would probably insist on limiting their reading to examinable content. That's fair; but after the exam, the interested reader is directed to this excellent essay by Nikolay Korolev (2021) which is literally titled "How potassium came to be the dominant biological cation". Without spoiling anything, the main thrust of the paper can be summarised as "good biochemical reasons, that's why". Primordial bacteria did not evolve to prefer potassium haphazardly; it was a choice based on the affinity of the ubiquitous peptide carbonyl oxygen atom for potassium over sodium. "The peptide bond is the most common part of all proteins, so its oxygen atoms become a natural choice for constructing the selectivity pore of the cation-conducting channels", Korolev speculates; and so it was just biochemically more convenient to make potassium-selective sodium-excluding channels. Intracellular potassium accumulation just happened to be the outcome of this. 

Mechanisms that regulate potassium homeostasis

Extracellular potassium is carefully regulated, as the gradient between the intracellular and extracellular potassium needs to be maintained as basically the highest priority your cells have. Some might say that a nicely separated potassium gradient is the very definition of life. Gumz et al (2015) discuss this subject with just the right lack of depth that a normal human can hope to understand it, whereas Palmer (2015)... 

Anyway. Homeostasis, for potassium, basically means making sure the intake and the output are matched to make sure the total body potassium stays reasonably stable, and using cells to hide any excess overflow. Intracellular potassium concentration can harmlessly fluctuate by something like 20 mmol/L in either direction, which makes the intracellular fluid a wonderfully flexible dumping ground for any excess potassium, or an emergency stash for when extracellular levels are low.

There are a couple of different ways that the regulation of potassium could be discussed. One is by covering the entrances and exits, so to speak. The benefit of this structure is that it hopefully leaves no molecule of potassium unaccounted for, and that hopefully means you mop up every regulatory mechanism and don's miss any exam marks. The basic structure proposed here is this:

• Regulation of potassium intake
• Regulation of the transcellular movement of potassium 
• Regulation of potassium elimination (renal and non-renal)

Regulation of potassium intake

To be completely frank, there is none. Some electrolytes, to be sure, have some regulated intestinal uptake mechanism (for example like magnesium and calcium), but this is not the case for potassium. Most sources say that 90% of dietary potassium absorption in the small intestine occurs by passive paracellular diffusion, which is completely unregulated. Absorption is purely driven by the concentration gradient. Dietary sources often have quite a substantial amount of potassium (for example, a half a cup of dried apricots has 1100 mg of potassium, which is approximately 30 mmol), so there's plenty of gradient to drive diffusion.  Demigné et al (2004) suggest that there must also be some kind of active absorption, in the sense that "it is not possible to exclude a component of active K⁺ transport, especially in the colon when dietary K⁺ intake is low". However, nobody knows what this is or how it works. For your own safety, do not mention this in your exam answers.

Regulation of transcellular potassium movement

The total amount of extracellular potassium is rather small. If you think about it, you only have about 13-14L of extracellular fluid, and therefore with 4mmol/L of potassium, you've probably only got 50mmol in total. Now, consider the abovementioned half a cup of dried apricots, with its 30mmol/L of potassium. If all of that potassium ends up dumped straight into the extracellular fluid, the potassium concentration would jump all the way to 5.7 mmol/L. Obviously, there's got to be some regulatory mechanisms that could be used to rapidly decrease the extracellular potassium concentration in these scenarios, to soften these peri-prandial concentration spikes. Sure, you could try to get rid of the excess potassium in the urine, but that's going to be a somewhat slower process - you've got to express some channels, move them to the apical membrane, then wait for some volume of urine to pass through those segments of tubule, all of which takes a while - say, twenty minutes at least before you see results. By that stage, your T-waves would be tented. Clearly a more rapid mechanism is called for. This is where the transcellular movement of potassium comes in.

Luckily, we have some idea of what the CICM examiners expected from this area of study. Question 4 from the first paper of 2012 asked for "the factors that affect the flux of potassium across the cell membrane".  Unlike most scenarios where the flux of something across a membrane is being asked about, this time it appears the examiners did not expect something brutally stupid like a rote-learned recitation of Fick's Law of diffusion. Or, at least, from a reading of their answer, one develops the impression that genuinely relevant factors were being asked about. Mechanisms governing the transcellular movement of potassium account for 90% of our medical strategies for the management of hyperkalemia, and it would have been a pity to waste that question on something like the Gibbs Donnan effect. In short, the following hormonal and environmental factors which determine potassium movements between the intracellular and extracellular compartments can be classified according to their molecular mechanism:

• Regulatory homeostatic influences:
  • Manipulation of Na+/K+ ATPase (eg. insulin, β-agonists)
  • Manipulation of ATP-sensitive K+ channels (glucose)
• Non-regulatory influences:
  • Activity of nonspecific cation channels (eg. NMJ)
  • Manipulation of extracellular osmolality
  • Manipulation of extracellular pH (bicarbonate)
  • Unexplained effects related to changes in temperature

Regulatory influences on transcellular potassium flux

You might call these "regulatory" because they are homeostatic in some way, i.e. they respond to a change in potassium in a way that reverses the change.  

Regulation of  Na⁺/K⁺ ATPase activity

Na⁺/K⁺ ATPase is the most effective lever to pull if you want potassium to get sequestered into cells in a hurry. This transmembrane pump is ubiquitous and susceptible to manipulation by multiple endocrine actors:

• Insulin increases the phosphorylation of the insulin receptor substrate protein (IRS-1), which leads to the activation of atypical protein kinase C and the insertion of extra Na⁺/K⁺ ATPase pumps into the membrane. This is the mechanism that is responsible for your ongoing survival after every time you eat a large potato. With food intake, insulin secretion also occurs, and potassium is safely sequestered inside cells.
• Aldosterone also increases the activity of Na⁺/K⁺ ATPase pumps in skeletal muscle (Phakdeekitcharoen et al, 2011). This probably falls into the overall spectrum of aldosterone activity, which is generally directed at removing or hiding extracellular potassium. Everything aldosterone does tends to have some anti-hyperkalemic effect. 
• β-agonists increase the activity of Na⁺/K⁺ ATPase pumps in skeletal muscle.  Specifically, this is a β2 effect. For the interested reader, Moratinos & Reverte (1993) expand on this in detail. The reason for this is thought to be the need to increase the availability of these transmembrane pumps to keep up with the demand of exercising muscle (a quantitatively much larger volume of ions will be in traffic across the muscle cell membrane if it keeps depolarising and repolarising all the time).

The attentive reader may wonder how this abuse of the Na⁺/K⁺ ATPase does not result in massive hypernatremia. Yes, the quantitative molar effect of this on sodium must be substantial, as for every two atoms of potassium pulled into the cell, three sodium atoms are pushed out. However, the effect is completely lost in the overall accounting of sodium. Observe: to drop your extracellular potassium from 7.0mmol/L to a safer 5.0mmol/L, your total extracellular fluid potassium needs to decrease from 98 mmol to 70 mmol. The total amount of potassium that enters the cells after the insulin/dextrose bolus is therefore only about 28 mmol.  Thus, the amount of sodium that is exchanged for this potassium must be 42 mmol. This extra 42mmol of sodium then distributes into an extracellular fluid compartment which already contains roughly 1,960 mmol of sodium.  The extracellular sodium concentration, therefore, increases from 140 mmol/L to 143 mmol/L. Most reasonable people would not raise an eyebrow at this change, as it is clinically insignificant.

Regulation of ATP-sensitive K+ channels

These are nucleotide sensitive cation channels on the membranes of numerous tissues, such as the vascular smooth muscle, cardiac muscle, skeletal muscle and pancreatic β cells (Rodrigo & Standen, 2005). Their role is to link membrane excitability to metabolism, i.e. they are open in the absence of ATP, decreasing the resting membrane potential and therefore preventing depolarisation to protect cells under stress conditions such as ischaemia or exercise. While they are open, they leak intracellular potassium into the extracellular fluid; therefore any influences which increase intracellular ATP will close these channels and reduce the extracellular potassium level. 

These factors include:

• Glucagon (decreases ATP, increases extracellular potassium)
• Glucose (increases ATP, drives extracellular potassium down)

Other substances can also interfere with their function:

• Sulfonylureas (directly close the ATP-sensitive K channels)
• Glibenclamide (directly closes  the ATP-sensitive K channels)
• Nicorandil (directly opens these channels, risking hyperkalemia)

Non-regulatory influences on transcellular potassium flux

These, as a group, might be described as "things which adjust the intracellular-extracellular potassium balance in a way which is not meant to act to address homeostasis, and is therefore not always constructive". They are potassium-rebalancing side effects of other physiological processes.

Potassium leak through nonspecific cation channels

Cation channels, such as the acetylcholine-gated sodium channels in the neuromuscular junction, are often nonselective. Activation of these channels allows sodium into the cell and potassium out of the cell. With scale, this flux could be sizeable. This is not a regulatory function of these channels, and under normal circumstances their effect on potassium goes unnoticed. Under some special circumstances, for example following stroke, spinal injury, or some other kind of denervation, these channels can proliferate on membrane surfaces, and their combined activity can become dangerous (eg. in the context of suxamethonium use).

Effect of acid-base disturbances on potassium

This is usually described as H⁺/K⁺ exchange, which is exactly what it is except with extra steps. It involves the Na⁺/H⁺ exchanger in skeletal muscle, which decreases in activity in states of acidosis. Aronson (2011) explains that this results in a deficit of intracellular sodium, which then decreases the substrate for Na⁺/K⁺ ATPase. So, now you have less potassium being pulled into the cell. At the same time, those stupid perpetually open transmembrane potassium leak channels are still leaking. Thus, when there is extracellular acidosis, not much potassium is being taken up into cells, and as the result acidosis and hyperkalemia tend to be associated. Reversing the acidosis increases the availability of intracellular sodium and allows the ATPase to entrain more potassium, lowering the extracellular potassium level. 

Effect of extracellular osmolality changes on potassium homeostasis

None of the official CICM sources that mention this mechanism support it with any references. However, with a bit of digging, it is possible to find if not the origin, then at least some acknowledgement of this mechanism in the literature. In short, there is objective evidence that extracellular hyperosmolarity causes hyperkalemia. Makoff et al (1971), for example, were able to raise the extracellular potassium level of dogs from by giving them mannitol. When Sotos et al (1964) increased the ECF osmolality of their rabbits to something like 500mOsm/kg with hypertonic saline, they observed the same thing (although this was quite extreme and some of the potassium might have come from cell lysis).  Conte et al (1990) trialled this in humans and noted an increase of about 0.6 mmol/L on average in response to about 6 mmol/kg of 5% saline (i.e. about 400 mmol for an average sized person).  In short, hyperosmolarity totally causes hyperkalemia.

What is the mechanism? Most resources concerned with the CICM primary suggest that intracellular volume contraction increases the intracellular concentration of potassium, and this increased concentration drives the efflux of potassium out of the dehydrated cell. This is also reproduced in Na (2005).  It makes a little bit of sense, because the constant leak of potassium out of the cell is happening because of its high intracellular concentration. However, the gradient is already between 5mmol/L and 140mmol/L, so how much higher could it get?  

Effect of temperature on potassium homeostasis

Temperature change produces a transcellular shift of potassium, but the mechanism underlying this is unclear. Hypothermia produces hypokalemia. Buse et al (2017) reviewed the literature and found crippling problems with all the current theories. All we know for sure is that it becomes intracellular, and this is completely reversed by rewarming.

Factors that influence transmembrane movement of ions in general

Yes, the Fick law is important. You should totally know it, you should be able to recite it when woken from a deep sleep by a distant gunshot. However, of the variables which determine Fickian diffusion (membrane thickness, surface area, diffusion coefficient, temperature, concentration gradient, the size of the potassium ion), literally none are under any direct control of the patient's own regulatory mechanisms, or susceptible to the manipulation of the clinician. The membrane will always be absurdly thin, and the total surface area (mainly referring to skeletal muscle) will always be absurdly vast; the potassium ion will always have the exact same predictable chemical properties, the concentration gradient will fluctuate within a very narrow physiologically tolerated range,  and the temperature will remain 37ish, as under most circumstances our patients are normothermic mammals (well...). Lastly, the diffusion coefficient for potassium ions in mammalian cell membranes is fairly stable (and surprisingly good) because these membranes are full of constantly open potassium leak channels.

The Nernst equation is also occasionally invoked to describe potassium equilibria across cell membranes. Again, yes it is an important relationship, as it describes the relationship of electrical and chemical forces acting on any ionic species across a semipermeable membrane. But all of the variables in that equation are stable and not under regulatory control. Moreover, the ion flux involved in the electrical events at the membrane is tiny. 

In summary, clearly passive Fickian diffusion or Nernst equilibria play basically no role in the actual regulation of potassium homeostasis; it just describes the playing field and the rules of the sport. If you want to have some control over your extracellular potassium, you only have facilitated transport mechanisms available.

Regulation of potassium elimination

The daily intake of potassium (usually about 70 mmol/day on a carby Western diet)  and the daily elimination of potassium must be well matched, so that you do not become some kind of concentrated potassium crystal. There are three main mechanisms by which the body can rid itself of unwanted potassium:

1. Rapid renal mechanism, in response to dietary intake
2. Slow renal mechanism, in response to changes in plasma biochemistry
3. Minor intestinal route of elimination

Aldosterone stimulates colonic ENaC channels to eliminate a small amount of potassium - this accounts for approximately 5% of the total daily potassium elimination, according to Batlle et al, 2015. So, one could quite reasonably omit it from the discussion, as it is a minor player. Many textbooks don't even mention it, which means it may not attract any marks in a written exam answer, as the tired hung-over examiner leafing through Ganong at the last minute to create their marking rubric will probably not see it. Renal handling of potassium is where the marks are, as it is the main mechanism of potassium regulation.

In summary:

• 50-60% of potassium is reabsorbed in the proximal tubule, which is not under any real regulatory control, and which mainly occurs by solute drag in the course of reabsorbing water
• 30% of filtered potassium is reabsorbed in the thick ascending limb due to the NKCC2 co-transporter, the drug target of frusemide; and again this is not thought to be under any physiologically important regulatory control
• Potassium is secreted into the tubular lumen in the distal convoluted tubule and the collecting duct, and this is the main lever of potassium regulation, controlled mainly by aldosterone.

Rapid kaliuresis in response to dietary potassium intake

 In response to an increase in dietary potassium, an increase in kaliuresis occurs prior to any systemic potassium level increase, which suggests that there is some sort of enteric or hepatic potassium sensor. When Calo et al (1995) gave 0.5 mmol/kg potassium citrate to their healthy volunteers, their urinary potassium output increased within twenty minutes. None of these people had any rise in their serum potassium level, and there was the most insignificant rise in aldosterone concentration. In fact adrenalectomised rats can do the exact same thing, which means aldosterone is not involved. In short, everything points to the existence of some sort of intestinal or portal venous potassium sensor. At this stage, nobody knows what this is, or how it works. 

Aldosterone-mediated renal elimination of potassium

Aldosterone increases the activity of ENaC channels, thereby stimulating the secretion of potassium (Young, 1988). The mechanism of this is related to the reabsorption of sodium. Because sodium is reabsorbed, the outer apical membrane becomes more negatively charged, and as the result more potassium leaks out via ROMK channels.  As more sodium is reabsorbed in the collecting duct, so more potassium is dumped.

This underpins the mechanism of diuretic-mediated potassium loss, as basically any mechanism which increases the delivery of sodium to the distal nephron will produce a loss of potassium. There is always some sodium being delivered to this part of the tubule, and therefore there is always some obligate dailt potassium loss. The minimum urinary potassium concentration is about 5mmol/L.

Feedback-regulated renal potassium loss

You could almost title this section "aldosterone-unrelated mechanisms".

• High potassium intake: leads to the increased expression of ROMK channels. This is a gradual adaptation to a chronically elevated potassium, and takes time to develop (Frindt & Palmer, 2010)
• Dehydration: leads to the systemic release of vasopressin, which also increases the availability of ROMK channels on the apical surface of principal cells in the collecting duct, and NKCC2 co-transporter in the thick ascending limb.  Increased potassium loss in this case may be some kind of useful move, to maximise the elimination of potassium even as urine output decreases. Just vasopressin trying to be helpful, perhaps? 

These mechanisms at least give the impression of being regulatory, in the sense that they tend to occur in response to situations where an increased potassium loss would be beneficial. On the other hand, there are several processes which affect renal potassium elimination, and which are completely unrelated to regulation, or often actually oppose it.

Non-regulatory influences on renal potassium elimination

These can be broadly summarised as "things that happen to renal potassium elimination which are completely unrelated to potassium feedback mechanisms and which are often totally counterproductive". These are:

• The effect of increased sodium delivery to the distal nephron
• The effect of systemic acidosis on renal potassium handling

As mentioned in the paragraph above, anything that increases the availability of sodium for the ENaC channel will produce a compensatory increase in potassium secretion to maintain electroneutrality. This is not a regulatory mechanism, and is frequently counterproductive. For example, in the presence of existing hypokalemia, you might want to hang on to your potassium, but if furosemide increases the delivery of sodium to the distal nephron, you will definitely keep losing potassium. One of the reasons is that most mechanisms that cause an increase in the distal sodium delivery are also mechanisms that increase diuresis and therefore water loss, resulting in systemic hypovolemia and aldosterone release. The extra aldosterone will of course be totally counterproductive here, trying to cling to sodium at the expense of potassium. The net effect of this system is a worsening hypernatremia and hypokalemia.

Metabolic acidosis causes distal potassium secretion to decrease, probably because of the extreme sensitivity of sodium and potassium channels to intracellular pH. Another mechanism is the action of the apical H⁺/K⁺ ATPase which can actively reabsorb potassium in exchange for hydrogen ions:
This is mainly happening  in the intercalated cells of the collecting duct. As you can see, acidosis (leading to a more acidic tubular fluid) will increase the reabsorption of potassium in exchange for H⁺, as a part of the renal handling of acid-base balance. Sure, that helps get rid of the acid, but in states of acidosis the systemic extracellular potassium tends to be increased. Reabsorbing more potassium in the nephron under those conditions seems like a dick move.

Drug therapy for hyperkalemia

This has appeared multiple times in the exam, and gives every impression that it will keep appearing. "Mechanism of lowering potassium, dosing, time to onset and duration of action" were the expected elements of a good answer in previous  SAQs.

It is better to classify these agents by their mechanism of action:

• Dilution
  • Rehydration (i.e. IV fluids)
• Intracellular movement
  • Bicarbonate (esp. isotonic bicarbonate)
  • β-agonists
  • Insulin
• Removal via the urine
  • Furosemide and other diuretics
• Removal via the bowel
  • Cation exchange resin

Fortunately, the mechanisms for most of these have already been discussed above, and for the rest the mechanisms are boring. Dilution. i.e. increasing the extracellular fluid volume with some potassium-poor fluid, is a totally legitimate way of decreasing extracellular potassium, but it is rather ineffective unless diuresis is the result. Consider: if your potassium is 7.0mmol/L, and you add an extra 2000ml of fluid to the extracellular fluid pool, you will end up decreasing the potassium concentration only down to 6.5 mmol/L.