Contents and properties of intracellular fluid

This chapter is vaguely relevant to Section E(iv) of the 2017 CICM Primary Syllabus, which expect the exam candidate to "describe the composition ...of intracellular fluid". Because this is the section on cellular physiology, the focus here will be mainly on the micro level of organisation, and all discussion of the intracellular body fluid compartment will be left for the Body Fluids section. 

The fact that the college examiners have never focused their Eye on this topic lends an almost liberating air of pointlessness to the discussion of it, as both the exam candidates and the author are pleasantly aware that each is wasting the other's time. Rather than imbibe hard science and memorise examinable facts here, one may easily read on for purely medutainment reasons, or skip ahead to topics which attract more marks. 

In summary:

What are the valid peer-reviewed resources for this topic? Unfortunately, they are numerous. When one uses a search engine or the index section of a textbook to look up "intracellular fluid", there is inevitably an answer, a series of entries or pages or powerpoint slides which - though all vaguely similar in their content - differ in their quoted values, and offer nothing in the way of references. Many offer no useful information whatsoever. For instance, turning to the official college books on this subject (Ganong, p.2 of the 23rd edition, and Guyton & Hall, p. 4 of the 13th edition), one finds cytosolic electrolyte concentrations being discussed using terms like "large amounts". Even outside of the official bibliography, otherwise solid offerings like Molecular Biology of the Cell do not have straight answers. They can't even agree on what to call it (cytosol? Protoplasm? Primordialschlauch?)

Fortunately, there are still scientists out there who publish on this topic. Probably the best article is the 1999 piece by Katherine Luby-Phelps, which basically contains everything you could possibly need to answer any hypothetical future SAQs on this topic. Another excellent entry dealing primarily with the properties of the intracellular material is the short paper by Richard P. Sear (2005). If one is truly mad and has the time resources of a permanently contracted staff specialist (i.e. no urgent obligation to actually perform any useful work) one may instead explore Gilbert Ling's In Search of the Physical Basis of Life (1984), an 800-page book written by a man whose publications on cell physiology stretch into the 1950s. 

Volume of an intracellular space

How big is a cell? Obviously, that depends on the cell. A good example of an outlier is the Xenopus oocyte, the egg of an African clawed toad which is 1mm in diameter. When discussing humans, one usually refers to intracellular fluid as if it is a relatively homogeneous bucket, but in fact that fluid compartment is composed of something like 10¹⁴ miniscule compartments each of which has a slightly different volume and composition.

The heterogeneity of all these volumes is well summarised in this chapter of the BioNumbers database, where all manner of meticulously referenced information can be found. It is reproduced here with minimal modification, in case the Harvard servers ever crash.

So, it's a fairly broad range. Moreover, obviously not all of the contents of the cell is going to be occupied by "intracellular fluid", irrespective of what your definition of that term. For example, for the purposes of this chapter, the definition of intracellular fluid will be "intracellular contents which are not organelles",  purely because organelles are discussed in another chapter. What we are talking about, in that case, is the "grayish, viscid, slimy, semitransparrent and semifluid substance" which occupies the intracellular space between other structures (Harvey, 1937).

Depending on what sort of cell you are looking at, that space could be a very small proportion of the total volume. For instance, in the abovelisted omental adipocyte, of those 60 picolitres of total volume, the vast majority will be occupied by waterless fat. This can be proven experimentally: DiGirolamo & Owens (1976) were able to calculate that the water volume in rat adipocytes was about 5-7% of the total volume, i.e. 1.5-2 picolitres.

In short, we are looking at a very small volume. Why does that matter? Well.  The cell's internal liquid 1-2 picolitre liquid volume is the volume of distribution for soluble substances. The molecules of those substances, therefore, have a very short distance to travel before they encounter one another. The effect is to increase the rate of reactions, which is helpful because the total amount of reagent molecules for such a small volume is by necessity small. To borrow an example from Luby-Phelps (2000), if a cell has a total content of 1 nanomole of a protein, that means there are only 1000 copies of that protein present in the cell. Fortunately, with such a small volume to traverse, a binding molecule with even lowish affinity would be able to scrub up and adsorb the majority of the available substrate.  

Protein contents of the intracellular fluid

Ok, so the volume is small. What macromolecules are in it, and how many of them? Alice B. Fulton (1982) weighed in on this with probably the most lucid answer in the published literature. Basically, the protein content of cells is in the range 17 to 35% by weight, with most authors falling in the range of somewhere like 20-30g/100ml. Fulton quotes ancient texts (Loewy et al, 1969) to give the following values:

• Muscle cells:  23% protein by weight
• Erythrocytes: 35% protein by weight 
• Most other cells: 17% to 26% protein by weight 

The measurements are usually done by refractive index measurements which is a technique which does not usually discriminate between structural proteins (eg. those of which the cytoskeleton and organelles are composed of) and soluble proteins which make up the rest of the goo.

So, what's the point of this discussion? Well. This concentration of protein is quite high. It is above the usually accepted concentration of large polymers which would be expected to affect diffusion of other like polymers, i.e. the forest is too dense. Chang et al (1987) produced a mathematical model which predicts that for polymers 50kDa and above, that diffusion limit is around 130g/L, i.e any more than that and other polymers will not be able to diffuse easily through the solution. Sure, they were using polystyrene dissolved in benzene, but the fact remains. For comparison, when one fully crystallises a protein, one ends up with a  "solid" which is only 40% protein by weight 

In summary, the proteins in the intracellular fluid are packed together so tightly that the cytosol has to be described as a "crowded solution". The cartoonish diagram here (originally published by Goodsell in 1993, and subsequently reproduced by virtually anybody who ever wrote about the cytosol) demonstrates visually exactly how closely packed those bodies are. The drawing was approximated, using known sizes and shapes of molecules/ electon microscope, but pictures taken through the SEM (eg. by Bridgman & Reese, 1984) demonstrate that it correctly represents the messy microstructure of the cytosol. Looking at Xenopus oocytes at a magnification of ~ 80,000, a dense forest of filaments and granules becomes apparent. The picture here (a part of their Figure 6) was actually cleaned up a bit by cell lysis and detergent wash, to remove some of the protein load which otherwise obscured the finer structure. Arrows are pointing at filament Y and T-junctions.

Without detergent preparation, all the fine granular material packed in between these filaments becomes visible. The picture now resembles white noise (same authors).

Clearly, diffusion through this thicket is not going to be normal. Smaller solutes (eg. your sodium and potassium ions) have to navigate around these huge obstacles, taking the scenic route towards each other. From a practical standpoint, this should mean that any reaction which depends upon the rate of diffusion is going to be slower. If molecules take forever to reach each other, surely the net rate of their interaction should be reduced. However, we do not see this.

What chemical properties do we see, with this highly saturated proteinaceous soup? Allen P. Minton (2006) has summarised years of (mainly his own) research in a table of the above-referenced paper. It is summarised here:

• Increased binding affinity of otherwise dilute macromolecules for one another
• Acceleration of protein associations eg. self-assembly
• Deceleration of diffusion-limited reactions and protein associations
• Stabilisation of proteins against heat denaturation

In summary, crowding forces proteins to fold and interact, which produces complex configurations which would be otherwise impossible in a dilute solution. For instance, Wilf & Minton in 1981 discovered that diluted myoglobin molecules in solution have little interest in one another, but the addition of a 10% solution of (any!) other protein causes the myoglobin to spontaneously assemble into dimers. 

Properties of intracellular water

Even in a crystallised protein, only 40% of the mass is actual protein. The rest is solvent which occupies the spaces between packed protein molecules (they are not exactly rectangles, and do not stack neatly). When the solvent is water, half of it ends up being adsorbed onto the protein surface, but the rest can still be viewed as normal liquid water. In this thin film, the ions of the intracellular water are dissolved.

Clearly, with the bound adsorbed water, things are slightly different. For instance, the solvent properties of this water are not going to be exactly the same as those of "free" water. For one, it will probably have reduced chemical activity. As one might expect, the "ordered" state of water gives it several unusual colligative properties - for instance, Foster et al (1976) found that its freezing point is depressed. Moreover, there are going to be pockets of cytosol with increased activity (around hydrophilic protein structures) and decreased activity (around the hydrophobic ones).

How much of the water is held in this captive state by the proteins? That's somewhat difficult to say. Experiments by Ling et al (1993) suggest that inside cells, most of the water is "ordered" in this way, but most of the experiments reporting on this subject are somewhat affected by the fact that the raw living cells they are using have various homeostatic responses to the experimental conditions which confound the findings.

Generally, people try to establish this answer by osmotically dehydrating cells. Apply an osmotic pressure, they reason, and all the "mobile" water should come out of the cell. Then you measure the cell's water content, and anything left over must be "immobile", parked on the surface of protein molecules and unable to migrate in response to osmotic pressure. Cameron et al (1997) present the graph (stolen shamelessly and displayed here, to the left) where the water content of the remaining cells is plotted against an x-axis of increasing osmotic pressure. The line of pressure vs. remaining water, when extrapolated from experimental evidence and extended towards an "infinite" osmotic pressure, ended up crossing the y-axis at some non-zero point. Depending on what cells you were using, that ended up being somewhere between 30-90% of the total water content.

In fact, it seems that this water adsorbed onto proteins is the essential water in the cells, and all the "free" water is pointless ballast. Clegg (1981), rehydrating some dried brine shrimp cells, found that metabolic activity resumed and was relatively normal when the cells were restored with about 35% water by weight. i.e. about as much as is expected to "hydrate" all the macromolecules. In those shrimp, there was almost certainly no "bulk phase" free water present. Sure, they weren't trying to reproduce or synthesise RNA (that required hydration up to 70-80%) but their synthesis of amino acids and gas exchange went on relatively normally. This makes all the more remarkable the fact that when measured objectively, this thick 20-30% protein soup has a viscosity which closely resembles normal water (Luby-Phelps, 1994)

Intracellular fluid pH

Carter (1972) published a highly influential article on this, which is often cited in human physiology textbooks even though the author used muscles from the giant barnacle (Balanus nubilus) which he suspended in something called "barnacle Ringer's solution", a brine with 450 mmol/L of sodium and 518 mmol/L of chloride in it. One can hardly read on with any level of respect for the human generalisability of such data, but if one does one would discover the most important finding: that pH in cells is compartmentalised so much that different regions of the cytosol had wildly different pH values, in a range of 6.0 to 7.5.   

Intracellular electrolyte concentrations

Generally, all textbooks, when asked about the concentration of electrolytes in the intracellular fluid, will produce a Gamblegram like this (misappropriated from Ling, 1984, with no permission from his estate or his publisher).

This diagram here is not referenced in Ling's book, but one might argue that it does not have to be, considering that Ling did virtually all the groundbreaking work on determining the behaviour of intracellular solutes. Specifically, in the 1960s, he and Ochsenfeld determined that this potassium (and all the other electrolytes in the cytosol) is generally not present in a freely available form, but is instead adsorbed onto the macromolecular structures. 

The investigators challenged cells with radiolabelled isotopes and found that this had little effect on displacing the electrolytes already present there, which would have been expected if they had been freely distributed. The intracellular electrolytes are largely present in complexes with macromolecules and are much less mobile then one might expect from a model of a cell where all the contents exist in a homogeneous watery sack. The same authors (Ling & Ochsenfeld, 1973) later confirmed that the mobility of potassium from the intracellular to the extracellular compartment was approximately one-eighth of what might be expected from simple diffusion in free solution. Killed frog muscle leaked potassium somewhat more readily (the diffusion rate was only reduced by 25% from the expected), as ATP-powered pumps stopped working and protein structure became disorganised.

So, let's agree these ions are not dissolved in a sloshing lake of free intracellular water, but rather bound in complexes. That still doesn't answer the question: how much of what is in there? Turns out, that's relatively easy to answer. Much easier, in fact, than any of the other questions brought up so far in this chapter. One merely needs to take one's cytosol, freeze-dry it, and then measure the elemental composition of the dry mass.  Mason et al (1981) did exactly this for some renal tubular cells, pre and post-ischaemic injury. Their findings are reproduced below, both in the form of the original 1981 table and a nice shiny Gamblegram:

As one can see from the wild fluctuations of potassium concentration following even 20 minutes of ischaemia, the electrolyte composition of cells is much more fluid than that of the extracellular fluid (where a 20mmol change in the concentration of any electrolyte would be poorly tolerated, at an existential organism survival sort of level). On top of that, every cell lineage will have some slightly different intracellular ion concentration. This gives rise to the imprecision of the intracellular electrolyte values discussed in textbooks, and their general reluctance to quote any numbers. Virtually any number you quote will be wrong. For example, Mason's distal tubules had 11mmol/L of sodium in their cytosol, but Poole-Wilson (1975) found about 44mmol/L in left ventricular myocytes and 20mmol/L in the left quadriceps. Alam et al (1977) give values of about 25 mmol/L for sodium and 145 mmol/L for potassium in some failing liver cells. In short, the messy and unpredictable environment of any given cell makes it difficult to quote specific numbers.