Brought to you by Portland Press Ltd.
Published on behalf of the International Federation for Cell Biology
Cancer Cell death Cell cycle Cytoskeleton Exo/endocytosis Differentiation Division Organelles Signalling Stem cells Trafficking
Cell Biology International (2012) 36, 503–509 (Printed in Great Britain)
Maintenance of low sodium and high potassium levels in cells and in tendon/collagen
Ivan L. Cameron*1, Anthony C. Lanctot† and Gary D. Fullerton†
*Department of Cellular and Structural Biology, University of Texas Health Science Center, San Antonio, TX 782293900, U.S.A., and †Department of Radiology University of Colorado Denver, 12700 East 19th Avenue, Aurora, CO 800452507, U.S.A.

Mammalian cells have a higher concentration of potassium and a lower concentration of sodium than their extracellular environment. The mechanisms responsible for the unequal distribution of these ions are commonly ascribed to the presence of an energy requiring plasma membrane ATPase pump, and the presence of membrane channels that pass one ion selectively, while excluding others. This report deals with other mechanisms that might explain this heterogeneous distribution of ions. To study other mechanisms, we turned to a non-living system, specifically tendon/collagen to eliminate the contribution of the membrane pump and channels. A simple gravimetric method was designed to measure solute accumulation or exclusion during rehydration of a well-washed, carefully dried and well-characterized protein specimen (tendon/collagen). Exposure to physiological salt concentrations resulted in selective exclusion of Na+ over K+, whereas exposure to low-salt concentration resulted in accumulation of these solutes. It is postulated that this solute redistribution occurs in all hydrated proteins and is partially responsible for the heterogeneous solute distribution in cells presently assigned to pump and channel mechanisms. Physical and thermodynamic mechanisms are offered to explain the observed heterogeneous solute distributions.

Key words: ion distribution, solute distribution, swelling water, tendon/collagen, water-of-hydration, water solvency

Abbreviations: SHM, stoichiometric hydration model

1To whom correspondence should be addressed (email

1. Introduction

1.1. Background to the low-sodium and high-potassium levels in cells

K+ concentration is relatively higher inside the mammalian cells and Na+ concentration is relatively higher outside mammalian cells (Table 1). This same unequal Na+ and K+ distribution apparently holds for nearly all vertebrate cells (Cameron et al., 1988).

Table 1 Concentration of sodium (Na+) and potassium (K+) ions in mM in vertebrate fresh water and terrestrial animal cells*

Intracellular ion concentration
Tissue [Na+] [K+]
Muscle cell
    Fish 28 138
    Amphibian 11 126
    Reptile 45 122
    Bird 21 83
    Mammal 8 135
Other cells
    RBC 15 143
    L cells 9 167
    Ehrlich cells 26 134
    Means±S.E. 20.4 137.3
    Plasma of five mammalian species 149 5
    Ratio: cell/plasma 0.137 27.5

Summarized from Table 2 in Cameron et al. (1988).

In inquiring about the mechanism responsible for unequal ion distribution, text books attribute it to the presence of a semi-permeable plasma membrane at the cell surface that contains: (i) a channel that allows passing of K+ but blocks Na+ and (ii) an energy requiring Na+/K+ ATPase exchange pump that pumps K+ in and Na+ out of the cell. The existence of a Na+/K+ pump has been documented beyond question. Likewise the patch-clamp technique that uses a micropipette tip to suction off a patch of cell membrane that remains attached to the micropipette tip has been demonstrating the opening of a single ion channel. This latter technique involves placing a steady bias voltage across a cell membrane patch and recording the current flow through the patch. The results reveal quantal size pulses, each of which is assumed to be due to the opening of a single ion channel. Questions arise about the ability of the Na+/K+ ATPase plasma membrane pump and the ion selective plasma membrane channel to explain adequately Na+ and K+ distribution. One question concerns the energetic ability of the pumping mechanism to maintain on its own the unequal ion concentration distribution, let alone operate all the other known membrane pumps requiring such enzymes (Ling, 1984). Another question concerns the interpretation of the patch clamp ion channel results, which is based on the findings that silicon rubber and other materials that definitely lack channels give similar results (Pollack, 2001). The size selectivity of channels in cell membranes has also been questioned (Pollack, 2001).

If the plasma membrane Na+ and K+ active transport pump and selective ion excluding membrane channel are not the sole determinates of ion distribution, then what else is involved to explain the heterogeneous ion distribution? Two possible mechanisms are: (i) the exclusion of sodium from the water-of-hydration associated with cytomatrix proteins and (ii) selective accumulation of potassium in the protein crowded in the cell internum.

The extent of water with non-bulk-like physical properties has been reported in mammalian lens, erythrocytes, skeletal muscle cells, sea urchin eggs, frog skeletal muscle and toad oocytes, as measured by water motion (proton NMR relaxation), flow rate (centrifugal dehydration force) and osmotic behaviour (summarized in Cameron et al., 2011; Cameron and Fullerton, 2006, 2008, 2011). Table 6 in the last report indicates that over half to almost all of cellular water has non-bulk-like physical properties. The nature of water in this environment could account for differing osmotic activities for K+ and Na+ (Fullerton, 2011).

Given that a majority of cell water differs from normal liquid water in its physical properties (slower motion, osmotic unresponsiveness, slower flow rate and higher viscosity), its solvent properties are worth exploring in greater detail. Molecules such as hydrated Na+ or sucrose in the bathing solution of a cell must first enter through cell membrane and work their way through non-bulk water. The energy needed for this diffusion process is greater here than in moving the same molecules through normal liquid water. The energy needed to move a larger solute molecule against a smaller solute molecule would be even greater. Also the larger the solute molecule, the greater is its rotational entropy, especially in non-bulk water. Finally, the ability of the solute molecule to fit into the structure of its surrounding non-bulk cell water needs to be considered.

A testable prediction of this theory is that a living cell will demonstrate solute exclusion as a function of solute size (molecular mass). Ling (2006) found that this theory holds for frog skeletal muscle cell. Given that the diameter of the hydrated Na+ is about 1.6 times that of the hydrated K+, selective exclusion of Na+ from at least some of the cells non-bulk water is predicted by solute size (Ling, 2006). Experimental support for the selective exclusion of Na+ over K+ also comes from the results obtained with a ‘cut-end' frog muscle preparation (Ling, 2001). Sodium ions in the bathing solution are excluded from the cytoplasm, whereas K+ accumulates in high concentration. Transmission electron microscopic images of the cut-end of frog muscles at various times after muscles transection showed no evidence of membrane formation at the cut surface (Cameron, 1988). The cut-end muscle preparation maintained its sodium exclusion for days. Even a gelatin gel excludes sodium from the bathing solution (Negendank, 1982, 1988). These observations support the idea that the low-sodium concentration in cells can be explained, at least in part, on a size-dependent cytoplasmic matrix exclusion mechanism.

Evidence for potassium accumulation by the cell's protein rich interior over the concentration of potassium in the bathing solution is strong. Cells that have their plasma membrane disrupted by exposure to mild detergents do not rapidly lose potassium into the bathing solution even though large holes in the cells plasma membrane could be seen by transmission and scanning electron microscopy (Kellermayer et al., 1986, 1994; Cameron et al. 1991; Hazlewood and Kellermayer, 1988). Edelmann (1988) demonstrated the strong selective accumulation of K+ over Na+ in the protein-rich A-brand of skeletal muscle against the protein poorer I-band. This finding is the opposite of that expected if potassium were dissolved in bulk water. This is because the protein-rich A-band has less, not more, potassium solvent water. Cameron and Hunter (1986) found that only 20% of a toad oocyte nuclear K+ is exchangeable (where K+ was replaced by Rb+ in the Ringers bathing solution), over the first 3–6 h of bathing, but no further exchange was observed after 40 h. This shows that the majority of intranuclear K+ is not free to diffuse from the oocyte nucleus. Other examples that cell potassium cannot freely diffuse in the cell can be found in Hazlewood and Kellermayer (1988), Cameron et al. (1988, 1996) and Ling (1984, 2001, 2006).

1.2. Choice of tendon/collagen as a biological material to help understand alternate mechanisms of ion distribution in cells

Because of the contribution of the plasma membrane pump and ion selective plasma membrane channels in living cells, it is not possible to test directly the alternate role of the selective ion exclusion by non-bulk water in the protein-rich interior of a cell or the role of the selective accumulation of K. Based on simultaneous operation of multiple ion distribution mechanisms in living cells, experiments were designed to study solute distribution in a membraneless bovine tendon/collagen specimen, which has been well characterized at the molecular level with almost pure aligned tightly packed type 1 collagen molecules. Native tendon has narrow water channels of ∼6 Å between adjacent collagen molecules (Fullerton and Amurao, 2006; Fullerton et al., 2010), appropriate for testing the hypothesis of ion exclusion dependence on hydrated ionic radii relative to the radius of water molecules.

Such a tendon specimen can be washed free of water soluble solutes and has low cellular content (2.6% tendon volume, see Materials and Methods section). Thus careful drying of such a washed tendon specimen yields the dry mass of a well characterized and pure molecular specimen free of water-soluble solutes. After weighing the washed and dried specimen to determine dry mass, it can be rehydrated in aqueous solutions of any water-soluble solute. By knowing the extent of the equilibrium weight of the bathed tendon specimen, one can get the total weight increase and by redrying and reweighing the specimen and calculate the extent of solute remaining in the dry tendon along with the amount of water taken up. The value can be expressed in g water or g solute per g of the initial dry mass or in terms of molar concentration of the solute in the water inside the tendon.

Data obtained by this simple gravimetric method can give information on both the selective solute exclusion of the tendon swelling water and solute accumulation by the tendon/collagen. This new information should provide further insight into the role of such alternate ion distribution mechanisms in cells.

2. Methods and materials

2.1. Tendon specimens

All experiments used bovine flexor tendons obtained at a local slaughterhouse from an animal ∼20 months old. Tendons were dissected in the laboratory and evaluated as either native tendon or dialysed to remove all mobile co-solutes. Tendon pieces were 1 cm in length.

2.2. Specimen preparation and characterization

Minimal variables would be preferable in studying water-of-hydration fractions in tendon/collagen. This section deals with two possible tendon non-collagen variables: (i) how much tendon cells contribute to the tendon/collagen volume fraction, (ii) can the water soluble co-solutes in native tendon be minimized or eliminated prior to initiation of the rehydration study.

Native bovine flexor tendon contains a fraction of tendon cells. Their volume fraction was determined by fixation of the fresh tendon with neutral-buffered formalin, followed by routine 7 μm histological H&E stained section preparations that were examined morphometrically in a light microscope with a ×100 oil immersion objective lens and an ocular grid (Weibel, 1969). The percentage of grid intercepts observed over cellular elements on bovine Achilles flexor tendons revealed a mean cellular volume fraction of just 2.3±0.65%.

To address the second issue, a measure of the initial dry mass of fresh native tendon was followed by extensive washing in deionized water, and careful drying and weighing of the washed tendon. This procedure allowed determination of water soluble material (solutes), in the fresh native tendon. The exact experimental conditions were to dry fresh tendon obtained within 3 h of slaughter either without or with washing in excess of deionized water for 5–7 days, with constant stirring and daily change of water. The tendon pieces were dried in a vacuum at room temperature for 1–2 days. The tendon pieces were heated in a vacuum oven at 90°C until weight equilibrium was achieved in 4–5 days. The unwashed and washed tendon pieces were weighed in the dried state. The differential between washed and unwashed samples allowed calculation of the extractable small molecule content. The mean amount of dry solute mass that could be washed from eight fresh tendon specimens was 3.7% of the unwashed fresh tendon dry mass.

Fresh native flexor tendon, when dried to weight equilibrium without washing, was light brown in colour; however, this same specimen dried was white when extensively washed in deionized water. The brown colour might have been due to glucose. Exposure of the well washed and dried white tendon to a 0.2 M glucose solution followed by re-drying turned to dark brown colour. When the glucose-exposed tendons were washed extensively in deionized water and dried they lost the brown colour. These findings indicate that the deionized water wash procedure efficiently removed a soluble brown colouring substance, e.g. glucose, from fresh native tendon.

To test whether the deionized water washing of fresh native flexor tendon removed detectable levels of tendon glucose, the tendon was cut into 1 mm pieces. The pieces were exposed to a known volume of deionized water (1 ml of water per 1 g wet tendon). This mixture was maintained at 4°C with continuous stirring for 6 days. A clinical glucose meter detected glucose in the wash water. The tendon pieces were blotted and rinsed in water three times before being homogenized with a mortar and pestle. No glucose was detected in the homogenate by the glucose meter. Thus the deionized water wash procedure extracted detectable levels of glucose from the flexor tendon. These observations demonstrate that washing of the fresh native tendon in deionized water reduces the amount of free glucose in the tendon to a level below that detected by brown colouration and by the clinical diabetic glucose meter. This finding does not, however, rule out the possible presence of bound glucose in the washed tendon that remained undetected by the measurement indicators used. Unpublished findings from this laboratory indicate the presence of residual glucose in tendon that was well washed in deionized water. It seems possible that this glucose may be in a form neither easily extracted from the flexor tendon nor detectable by brown colouration.

The above observations all indicate that fresh native tendon is low in cellular volume fraction and the well-washed dried tendon has low levels of mobile solute in the dry-weight fraction. Thus one possible variable, cellular volume fraction, has been quantified and one variable (water soluble solutes) has been minimized or eliminated. The well washed and carefully dried tendon was therefore judged to be a better model specimen than fresh native tendon for study of the solvent properties of tendon/collagen water-of-hydration.

2.3. Dialysis method

It was shown as described above that all measurable mobile sugars and ions can be removed by exposure of tendon samples at room temperature to an excess of pure distilled water for 5–7 days with constant stirring and daily change of water.

2.4. Calculation of hydration mass

As previously found, tendon can be completely dried without damage to the proteins by first drying in a vacuum at room temperature for 1–2 days. The residual bound water of ∼0.26 g water/g dry weight following room temperature drying was removed by heating the tendon pieces in a vacuum oven to 90°C until a second weight equilibrium was obtained after 4–5 days. The vacuum oven was backfilled with dry nitrogen and the samples weighed immediately to prevent rehydration from the laboratory air. The final dry mass was assumed to be the dry mass of dry protein and any remaining co-solutes. The mass of hydration water was calculated by the difference between the initial wet mass and the final dry mass.

2.5. Solute and water uptake measurements

Dried and weighed tendon pieces, following initial dialysis, were placed in bathing solutions at 22°C for 1–7 days depending on the sample size until the blotted weight of the tendon reached weight equilibrium. Ten different bathing solutions were used: (i) distilled water, (ii) Dulbecco's PBS, (Dulbecco and Vogt, 1957), (iii) 0.065 M NaCl, (iv) 0.125 M NaCl, (v) 4.64 M NaCl, (vi) 0.076 M KCl, (vii) 0.125 M KCl, (viii) 4.67 M KCl, (ix) 1.14 M NaSO4 and (x) 0.186 M glucose. The mass of the water uptake was measured, and the mass of solute uptake was measured as the difference of the final dry mass following solute exposure minus the final dry mass following dialysis in pure water.

3. Results

3.1. Extent of tendon hydration (swelling) in various bathing solutions

We determined the maximum level of swelling of deionized water washed and dried bovine flexor tendons suspended in various solutions. The data in Table 2 lists the weight equilibrium water uptake values. For reference, fresh native tendon had 1.62 g water/g dry mass (g/g), which matches the water monolayer value of 1.6 g/g predicted by SHM (stoichiometric hydration model; Fullerton and Cameron, 2007), but the water washed and dried tendon, swelled to only 1.2 g/g when placed in deionized water. When washed and dried tendon was hydrated in 0.065 M NaCl or in 0.076 M KCl, hydration increased to near that of fresh native tendon. However, when the washed and dried tendon was exposed to near saturation concentrations of NaCl or KCl (4.6 M), the extent of hydration was reduced to 0.69 g/g for NaCl and 0.77 g/g for KCl. It is noteworthy that the tendon swelling for the concentrated KCl solution against the concentrated NaCl solution was slightly and significantly higher (Table 2). Exposure of the washed and dried tendon to a 0.186 M glucose solution led to hydration to about the same level (1.52 g/g) as the two lowest concentration of salts (i.e. 1.58 and 1.59 g/g). Again it is noteworthy that the two lower concentrations of salts also allowed swelling of the tendon pieces to a level near to that of fresh tendon (1.62 g/g).

Table 2 Equilibrium hydration level in deionized water washed and dried samples of bovine flexor tendon that were then exposed to various solutions

Grams of water per grams of dry mass (g/g)±S.D., n = 3 or 4.

Solution exposure condition* g/g±S.D.
Deionized water 1.20±0.016
Deionized water+0.065 M NaCl 1.58±0.044
Deionized water+0.125 M NaCl 1.414±0.064
Deionized water+4.64 M NaCl 0.690±0.014
Deionized water+0.077 M KCl 1.59±0.010
Deionized water+0.125 M KCl 1.463±0.047
Deionized water+4.67 M KCl 0.767±0.004
Deionized water+1.14 M NaSO4 0.693±0.011
Deionized water+0.186 M Glucose 1.52±0.050
Other samples
    Fresh native tendon 1.62±0.035
    Fresh native tendon stored in PBS+0.5% sodium azide 2.23±0.057

In all solution exposure condition only the NaCl of 4.64 M versus the NaSO4 of 1.14 M, the NaCl of 0.069 M versus the KCl of 0.077 M and the NaCl of 0.125 versus KCl of 0.125 are not significantly different. An ANOVA statistical test followed by an SNK multiple range test was used to test for significant differences.

When fresh native tendon was stored in Dulbecco's PBS with 0.5% sodium azide the native tendons that had not been washed prior to exposure to PBS, swelled to 2.23 g/g. This is a value significantly above that observed when the washed and dried tendon was brought to weight equilibrium in 0.068 M NaCl or in 0.076 M KCl (1.55 or 1.59 g/g) and to a level significantly higher than fresh native tendon (1.6 g/g).

As indicated in Table 2, the three bathing salt solutions ranging from 1.14 to 4.47 M when added to washed and dried tendon/collagen caused swelling to hydration levels of 0.69 and 0.77 g/g. This level of hydration is near the expected mean level of hydration for water coverage over the native collagen polar surface area as predicted by the SHM (Fullerton and Cameron, 2007; Cameron et al., 2011), and is consistent with the level of hydration where the hydration force necessary to remove water increases precipitously (Leikin et al., 1994, 1997; Fullerton and Rahal, 2007). Data in Table 2 also indicated that exposure of washed and dried tendon to relatively low concentrations of NaCl (0.068 M) or KCl (0.065 M) allowed swelling to the hydration level of fresh native tendon. Thus the 2 lower salt concentration solutions as well as the 0.186 M glucose bathing solution allowed hydration to the level of the native tendon and to the level of water monolayer coverage over the tendon/collagen surfaces, as predicted by the SHM (Fullerton and Cameron, 2007, Cameron et al., 2011). It should be noted, however, that rehydration of the washed and dried tendon to the deionized water allowed swelling to only 1.20 g/g, which is water content below that of a monolayer of water predicted by the SHM. Thus, the presence of some salt or glucose was necessary to get rehydration to the level of the water content of fresh native tendon and the level of a monolayer of water coverage as predicted by the SHM.

3.2. Solute exclusion and accumulation upon rehydration of well washed and carefully dried tendon in different aqueous bathing solutions

Knowing (i) initial dry mass of the well washed and carefully dried tendon, (ii) weight increase following exposure to the specific bathing solutions of known chemical composition and (iii) final dry mass of the swollen tendon, one can calculate solute concentration in the aqueous environment within the tendon after the swelling process reaches weight equilibrium. The extent of swelling (in g/g), is given in Table 2. The table also lists the concentration of the outside bathing solution, and Table 3 lists the calculated concentration of solutes in the aqueous environment both outside and inside the swollen tendons. The ratio of the concentration of solutes inside the tendon over the concentration outside the swollen tendon is also listed in Table 3. In some of the bathing solutions, the concentration in the swollen tendon was greater than the concentration of the bathing solution, and in some cases the concentration in the swollen tendon was less than the bathing solution. Inside concentration lower than the outside concentration demonstrates solute exclusion from the tendon swelling water.

Table 3 Solute exclusion and accumulation upon rehydration of well washed and dried tendon in different bathing solutions (means±S.D., n = 3–4), mol/l.

Bathing solution Concentration outside Concentration inside Ratio concentration inside/outside*
KCl 4.67 2.92±0.071 0.623±0.015
NaCl 4.64 1.10±0.269 0.237±0.058
KCl 0.125 0.104±0.011 0.832±0.066
NaCl 0.125 0.092±0.007 0.614±0.088
KCl 0.077 0.305±0.023 4.013±0.307
NaCl 0.065 0.384±0.039 5.647±0.560
NaSO4 1.14 0.508±0.017 0.446±0.047
Glucose 0.186 0.368±0.009 1.977±0.043
Mammalian cells in vivo†
Na+ 0.149 0.02 0.13
K+ 0.005 0.137 27.4

A ratio <1 demonstrates solute exclusion and a ratio >1 demonstrates solute accumulation. All tendon ratios listed in this table are statistically different except for the ratio of the two lowest outside concentration salt bathing solution conditions and between the high concentration of KCl and NaSO4.

See Table 1.

The five bathing solutions with higher molarities (NaSO4 1.14 M, NaCl 0.125, KCl 0.125, 4.67 M, NaCl 4.64 M) all demonstrate salt exclusion from the tendon water. Although the outside concentration of KCl and of NaCl is almost the same the ratio of inside over outside concentration is significantly lower for NaCl than for KCl. This indicates that NaCl is selectively excluded from the tendon swelling water against the KCl, and this selective exclusion of NaCl occurred at physiological levels of salt in the bathing solution.

Another factor besides solute exclusion from water-of-hydration, comes into play in a study of solute distribution in dried tendon rehydrated in different bathing solutions. This is the possibility that the solute in the bathing solution can accumulate in the tendon/collagen. This is indeed what was found when dried tendon was exposed to bathing solutions with relatively low concentrations of KCl, NaCl and glucose (Table 3). When monovalent salt concentration is converted into particle concentration by assuming perfect dissociation into 2 particles, one anion and one cation as assumed in Figure 1, the uptake of particles depends only on their number in solution. This is clearly a colligative property governed by osmotic relationships. As the gravimetric method does not allow one to follow the accumulation of the two individual cations (Na+ and K+) when mixed in the bathing solution, the selectivity of the tendon accumulation of one over the other ion cannot be directly determined by this gravimetric method.

4. Discussion

By studying washed and dried tendon, we are concerned with a well-organized macromolecular protein structure of almost 100% pure aligned type 1 collagen free of water soluble solutes. The separation of molecular surfaces simulates water in the crevices and interstices between globular and fibrillar proteins in the cytoplasm of cells. Water absorbed by carefully washed and dried tendon specimens can be considered water-of-rehydration and the fact that the water sorption rate reveals four hydration fraction indicates the presence of multiple water-binding sites of different potentials (Cameron et al., 2011) and solvation properties (as indicated above). A tendon/collagen molecular SHM has been described that explains the size and biophysical properties of each of the four water-of-hydration sites (Fullerton and Cameron, 2007). This SHM appears applicable to globular proteins and to cell systems where it has been tested (Cameron et al., 2011).

There has been considerable debate about the solvent properties of the water-of-hydration associated with proteins and other materials. On the one hand, it is clear that ice excludes ions and other co-solutes. When placed in a dye solution ice remains clear, thus water in the ice cube is a non-solvent for the dye. On the other hand, Hill (1930) demonstrated that water in frog muscle comes to the same concentration of urea as that in the bathing solution, which led him to conclude that all water in the cell has bulk properties for all solutes. Ling has since repeated this study using urea and other solutes (Ling, 2001) and confirmed the urea result, but found at least partial solute exclusion with many other solutes (Ling, 2001), thus demonstrating that not all water in a cell has solvent properties as bulk water.

Expressing concentrations inside the tendon, as calculated and presented in Table 3, is a combination of two factors. The amount of solute accumulated in the tendon/collagen at low concentrations and the ion or molecule specific amount of solute excluded from the tendon is most visible at high co-solute concentrations. This leads to the conclusion that solute accumulation does occur when the bathing solution has the lower concentration of solutes, and that solute exclusion does occur in the water-of-hydration when the concentration of the solutes in the bathing solution is at a much higher concentration.

Table 3 shows that there is significant evidence of accumulation of solutes to the two lowest concentrations of KCl and NaCl in the bathing solution, but no significant evidence for an ion selective difference. One may reasonably ask if evidence for selective accumulation of KCl against NaCl can be found at even lower concentrations in the bathing solution (G.D. Fullerton, A.C. Lanctot and I.L. Cameron, unpublished work).

The data in Table 3 of the ratio of solute inside to that outside the tendon, when either KCl or NaCl was added to the bathing solution at nearly the same higher molar concentration, indicates that the sodium salt is selectively excluded from the water-of-hydration compared with the potassium salt. Notice, however, that the total extent of water-of-hydration sorbed by dry tendon bathed was 0.69 g/g in 4.6 M NaCl and 0.77 g/g in 4.6 M KCl. These levels of hydration come close to the level of water-of-hydration expected over the hydrophilic (polar) surface of tendon/collagen as predicted by the SHM (0.8 g/g). One may deduce from this that solute exclusion caused a concentration differential between internal and external water that creates an osmotic hydration force that collapses the gap between collagen molecules, as proven by the hydration force experiments of Leikin (1994, 1997).

A mechanism is proposed to explain the solute accumulation and exclusion results listed in Table 3. Figure 1 is a plot of water uptake by rehydrated tendon as a function of particle concentration for the ionic entities as well as glucose. Figure 2 gives a plot of the calculated molar ratio of the solutes inside the tendon over outside the tendon as a function of particle concentration (complete disassociation of salts is assumed). Two contrary effects that explain the results in Figures 1 and 2 are presented in Figure 3.

The first effect is the increased viscosity of the bound water fractions on tendon that slows diffusion of solutes relative to solutes in bulk water. The second effect is solute exclusion due to the larger size of solutes relative to the much smaller size of the water molecule.

At low particle concentrations, the slower motion of solutes in the tendon causes solutes to accumulate in the spaces between collagen molecules, and thus the concentration inside the tendon increases to several times the concentration in the mother solution. The increased concentrations of solute inside the tendon create an osmotic pressure differential for water that brings additional water into the space between collagen molecules thus increasing hydration from 1.2 g/g to values in excess of 2.3 g/g. This reduces the global viscosity of the tendon water-of-hydration and causes uptake of additional water-of-hydration.

At higher solute concentrations, the solute exclusion effect dominates and tips the osmotic relationship in the opposite direction (Figure 2). Exclusion of solutes creates an osmotic differential that compresses water from the space between collagen molecules, similar to the hydration force experiments of Leikin et al. (1994, 1997). This comes to equilibrium near hydration = 0.724 g/g, the SHM limit of polar surface hydration demonstrated by Leikin et al. (1994, 1997).

Data in Table 3 indicate selective exclusion of NaCl against KCl at high solute concentrations suggesting that particle size is involved given that the hydrated volume of the sodium ion is larger than the hydrated potassium volume.

In conclusion, the energy requiring membrane pumps and ion selective channels are not the only mechanisms involved in the heterogeneous solute distribution between a cell and its extracellular environment. The relative importance of other mechanisms discussed herein has yet to be established, but cannot be dismissed or ignored.

Author contribution

All authors contributed equally to the experiments and writing of the paper.


This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.


Cameron, IL and Hunter, KE (1986) Subcellular flux of potassium and rubidium in amphibian oocytes. Physiol Chem Phys Med NMR 18, 133-6
Medline   1st Citation  

Cameron, IL (1988) Ultrastructural observations on the transectioned end of frog skeletal muscle. Physiol Chem Phys Med NMR 20, 221-5
Medline   1st Citation  

Cameron, IL, Fullerton, GD and Smith, NKR (1988) Influence of cytomatrix proteins on water and on ions in cells. Scanning Microsc 2, 275-88
Medline   1st Citation   2nd   3rd  

Cameron, IL, Cox, LA, Liu, XR and Fullerton, GD (1991) Maintenance and mobility of hemoglobin and water within the human erythrocyte disruption of the plasma membrane. J Cell Physiol 149, 362-72
1st Citation  

Cameron, IL, Hardman, WE, Fullerton, GD, Miseta, A, Koszegi, T and Ludany, A (1996) Maintenance of ions, proteins and water in lens fiber cells before and after treatment with non-ironic detergents. Cell Biol Int 20, 127-37
Crossref   Medline   1st Citation  

Cameron, IL and Fullerton, GD (2006) Non-bulk-like water in cellular interfaces. In Water and the Cell (Pollack GH, Cameron, IL and Wheatley, DN, eds), pp. 315-24, Dordrecht, Springer
1st Citation  

Cameron, IL and Fullerton, GD (2008) Interfacial water compartments on tendon/collagen and in cells. In Phase Transition in Cells (Pollack GH and Chin, WC, eds), pp. 43-80, Springer, Berlin
1st Citation  

Cameron, IL and Fullerton, GD (2011) Lanctot AL. The molecular stoichiometric hydration model (SHM) applied to tendon/collagen, globular proteins and cells. Cell Biol Int 35, 1205-15
Crossref   Medline   1st Citation   2nd   3rd   4th   5th   6th  

Dulbecco, R and Vogt, M (1957) Plaque formation and isolation of pure lines with poliomyelitis virus. J Exp Med 106, 167-9
1st Citation  

Edelman, L (1988) The water problem posed by electron microscopic studies of ion binding in muscle. Scanning Microsc 2, 8851-65
1st Citation  

Fullerton, GD (2011) Differential measure of water structuring for ionic salt solutions. J Phys Chem
1st Citation  

Fullerton, GD and Cameron, IL (2007) Water compartments in cells. Methods Enzymol 428, 1-28
Medline   1st Citation   2nd   3rd   4th  

Fullerton, GD and Rahal, A (2007) Collagen structure: the source of the tendon magic angle effect. J Magn Reson Imaging 25, 345-61
Crossref   Medline   1st Citation  

Fullerton, GD, Amurao, MR, Rahal, A and Cameron, IL (2010) Micro CT dilatometry measures of molecular collagen hydration using bovine extensor tendon. Med Phys 37, 1-14
Crossref   Medline   1st Citation  

Fullerton, GD and Amurao, MR (2006) Evidence that collagen and tendon have monolayer water coverage in the native state. Cell Biol Int 30, 56-65
Crossref   Medline   1st Citation  

Hazlewood, CF and Kellermayer, M (1988) Ion and water retention by permeabilized cells. Scanning Microsc 2, 267-73
Medline   1st Citation   2nd  

Hill, AV (1930) The state of water in muscle and blood and the osmotic behaviour of muscle. Proc R Soc Lond B Biol Sci. 106, 477
1st Citation  

Kellermayer, M, Ludany, A, Miseta, A, Koszegi, T, Beta, G and Bogner, P (1994) Release of potassium, lipid, and protein from detergent treated chicken red blood cells. J Cell Physiol 159, 197-204
Crossref   Medline   1st Citation  

Kellermayer, M, Ludany, A, Jobst, K, Szucs, F, Trombitas, K and Hazlewood, CF (1986) Cocompartmentation of proteins and K+ within the living cell. Proc Natl Acad Sci USA 83, 1011-5
Crossref   Medline   1st Citation  

Leikin, S, Parsegian, VA, Walrafen, GE and Yang, W (1997) Raman spectral evidence for hydration forces between collagen triple helices. Proc Natl Acad Sci USA 94, 11312-7
Crossref   Medline   1st Citation   2nd   3rd   4th  

Leikin, S, Rau, D and Parsegian, VA (1994) Direct measurement of forces between self-assembled proteins: temperature-dependent exponential forces between collagen triple helices. Proc Natl Acad Sci USA 91, 276-80
Crossref   Medline   1st Citation   2nd   3rd   4th  

Ling, GN (2006) A convergence of experimental and theoretical breakthrough affirms the P M theory of dynamically structured call water on the theory's 40th birthday. In Water and the Cell (Pollack GH, Cameron, IL and Wheatley, DN, eds), pp. 1-52, Springer, Berlin
1st Citation   2nd   3rd  

Ling, GN (1984) In Search of the Physical Basis of Life, pp. 227-69, New York, Plenum Publishing Co
1st Citation   2nd  

Ling, GH (2001) Life at the Cell and Below-Cell Level., pp. 52-54, 92–101, Pacific Press, New York
1st Citation   2nd   3rd   4th  

Negendank, W (1982) Studies of ions and water in human lymphocyts. Biochim Biophys Acta 694, 123-61
Medline   1st Citation  

Negendank, W (1988) Overview of physical studies of bulk water in biopolymers. Scanning Microsc 2, 867-70
Medline   1st Citation  

Pollack, GH (2001) Cell Gels and the Engines of Life, pp. 11-6, Seattle, Ebner and Sons
1st Citation   2nd  

Weibel, ER (1969) Stereological principles for morphotometry in electron microscopic cytology. Int Rev Cytol 26, 235-302
Medline   1st Citation  

Received 8 August 2011/19 December 2011; accepted 1 February 2012

Published as Cell Biology International Immediate Publication 1 February 2012, doi:10.1042/CBI20110439

© The Author(s) Journal compilation © 2012 International Federation for Cell Biology

ISSN Print: 1065-6995
ISSN Electronic: 1095-8355
Published by Portland Press Limited on behalf of the International Federation for Cell Biology (IFCB)