Fluoridation: Aspects of toxicity

[Malcolm Harris, Ph.D. (Wales), B.Pharm (Wales), FPS, Fss, FRSH]

Fluoridation: Errors and Omissions in Experimental Trials

[Chapters 19, 20 and 21. Philip Sutton. Originally published in 1960]

The Kinetics of Acetylcholinest'se Inhibition and the Influence of Fluoride and Fluoride Complexes on the Permeability of Erythrocyte Membranes

[Westendorf, 1975]

Introduction | Contents | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9

The Kinetics of Acetylcholinesterase Inhibition and the Influence of Fluoride and Fluoride Complexes on the Permeability of Erythrocyte Membranes - Page 8.

2. Presentation of the Permeability Experiments  

We next carried out model experiments in which, with regard to the material being examined, we limited ourselves to red blood cells because red blood cells are very easy to obtain and require relatively small experimental effort to handle. If, for example, one wants to track the distribution of a radioactive material between serum and erythrocytes, one lets the test substance act on fresh whole blood, to which a material to prevent coagulation must be added during extraction, and afterwards separates the cells from the serum by centrifugation. After centrifuging one determines the radioactivity in both phases. One derives the distribution gradient by examining the volumetric ratio of the two phases. 

Of course the findings derived from these cells cannot simply be transferred to other cell systems without due consideration. Since the underlying principles for these processes are, however, the same throughout the body, valuable implications can be derived if, for example, one considers the function of the active transport of Na⁺ and K⁺ as well as the influence of fluoride on these processes. (Na⁺-K⁺) activated ATPases, for example, which are localized in the cell membranes and are tightly coupled with the cation transport, can not be distinguished from each other, no matter from which organ they originated(32). The same is true, although to a lesser extent, for the enzymes of the glycolysis chain.  

The erythrocytes' main assignment is the transport of O₂ and CO_2­ as well as the stabilization of the blood pH associated with this transport. Erythrocytes contain no sub cellular particles, but therefore large amounts of hemoglobin. They derive energy for active cation transport from glycolysis. The pentose-phosphate cycle supplies the cell with the redox catalyst NADPH. 

In terms of diffusion of substances through the cell membranes, we distinguish between passive and facilitated diffusion as well as active transport. Na⁺, K⁺, Ca²⁺, and Mg²⁺ fall into the latter category with certainty, glucose into the middle one. We first tried to determine the concentration range in which an effect of fluoride on the diffusion processes is noticeable.  

a. Potassium Permeability 

The potassium content of the erythrocytes is around 10 times as large as that of the serum. The permeability of the membrane for K⁺ is, however, small, but nonetheless clearly present. If one supercools the erythrocytes, a potassium equalization occurs between the cell and the serum, since the active transport mechanisms are inhibited. If the cells are afterwards warmed back to 37⁰C the original concentration gradient is reestablished. We made use of this behavior when labeling the erythrocytes with ⁴²K. First we produced a ⁴²K labeled Ringer's solution, which consisted of the following: 

NaCl                            0.9 g

CaCl₂ · 4H₂O              39.6 mg

NaHCO₃                      50 mg

Glucose                        50 mg

MgCl₂                          5.4 mg

KCl                              19.9 mg ~ 50µCi

------------------------------------------

double-distilled water 1,000ml

Next we withdrew blood from a lightly blocked arm vein (addition of 10% isotonic citrate solution) and separated the cells and the serum by centrifugation. After removing the serum and the leukocyte layer, which lies on the erythrocytes as a thin film, we washed the cells three times with isotonic NaCl solution and then suspended them in Ringer's solution (as above, however without ⁴²K). The proportion of the cellular volume (hematocrit value) was now 40.2%.  

We mixed 0.5ml of the suspension with 0.5ml of the radioactive Ringer's solution in each of 10 test tubes. The specimens were then stored for half an hour at 2^o C. After that they warmed at 37^o C, again for half an hour, during which time the re-exchange of K⁺ that had escaped in the cold was supposed to occur. We freed the cells treated in this way by centrifuging and washing two times to release the bound Ringer's solution, which we then replaced with an equal amount of non-labeled solution.  

Next we added two additional aliquots of 0.1ml of the solution and 0.1ml NaF solution to each of the remaining specimens, so that the final concentration became 5 x 10^-5 – 10^-2 M. After the specimens were held for half an hour longer at 37^o C, we extracted 0.25ml of supernatant from each after centrifugation, and determined the radioactivity in the liquid scintillation counter on the basis of "Cerenkov radiation" We hemolyzed one of the fluoride free specimens before the extraction by quickly dipping it in liquid oxygen in order to derive a value for the saturation after complete equalization of the ⁴²K. The following figure reproduces the course of the change in the activity of the solution as a function of the F^- concentration.  

Figure 33. ⁴²K Outflow From Erythrocytes as a Function of the NaF Concentration. 

Reaction time 0.5 hours. 

No effect can yet be observed at concentrations of less than 10^-4M NaF. If the peak at 2.5 x 10^-4M NaF is real or is only based on an error of measurement can not be determined from these data. K-42 concentration increases linearly between 0.5 and 5 x 10^-3M NaF, and then slowly changes over into saturation. 87% saturation is reached at 10^-2M NaF. So the K⁺ concentration of the cell begins to diminish at F^- concentrations above 10^-4 = 1.9 mg F/l. One can therefore assume that an increase in the flow of potassium out of the cell is possible at the upper boundaries of the physiological region. The reason for the fluoride dependent K⁺ efflux can not be determined from this measurement. 

b. Calcium Permeability 

Erythrocytes have a relatively small Ca²⁺ concentration in comparison to the serum. According to ROMERO and WHITTAM(31), Ca²⁺ takes on a regulatory role in the active transport of Na⁺ and K⁺. The Ca²⁺ is itself supposedly "pumped" out of the cell by an ATP dependent "Ca-pump". In the authors' opinion, fluoride can only influence the K⁺-Na⁺ permeability in the presence of Ca²⁺. Evidence has suggested that the calcium ions cause an inhibition of the (Na⁺-K⁺) activated ATPase by way of a competitive reaction with magnesium ions. (33) We again asked ourselves at which fluoride concentrations the intra-cellular concentration begins to change. 

We again used a suspension of erythrocytes in Ringer's solution, as described for the potassium. We doped a series of specimens of different NaF concentrations with an incalculable amount of ⁴⁵Ca (as CaCl₂) and incubated them for 18 hours at 37⁰C. The ⁴⁵Ca thereby distributed itself evenly over the intra- and extra-cellular spaces in correspondence with the natural Ca²⁺ gradient. When the time was up we separated the cells by centrifugation, washed them once with Ringer's solution, and after hemolysis determined the radioactivity of the cells using the liquid scintillation counter. To prevent the disappearance of color we precipitated the hemoglobin with trichloroacetic acid, whereby we assured ourselves that the precipitate did not contain any Ca-45, which was the case after rinsing once with 10% trichloroacetic acid. Figure 34 plots the ⁴⁵Ca content of the cells as a function of the NaF concentration. 

Figure 34 - Intra-cellular ⁴⁵Ca Concentration as a Function of the NaF Concentration. 

Reaction time 18 hours 

The plot shows great similarity to the respective potassium curve. Development of a maximum can again be observed at 5 x 10^-4M NaF, which is, however, significantly more distinct in this case than with the potassium. After 2.5 x 10^-3M NaF the Ca²⁺ concentration in the cell linearly approaches the fluoride concentration. This behavior speaks for a direct relationship between the Ca²⁺ and the F^- concentrations in the cell.   

Although the solubility product of CaF₂ is exceeded at F^- and Ca²⁺ concentrations > 2 x 10^-4M, precipitate formation most likely does not occur immediately, since a large proportion of the Ca²⁺ present is bound to proteins and other complex forming substances. On the other hand, it is highly conceivable that the Ca²⁺ is present with one valence bound to an organic molecule and the other bound to F^-. This binding pattern would also explain the linear relationship between of the Ca²⁺ and F^- concentrations.  

c. Sodium Permeability 

The cell membranes of most animal organisms, as well as those of many plants, have the ability to actively transport Na⁺. The movement of Na⁺ plays a role in, among others, excitation of nerve cells, maintenance of the cell potential, protection of erythrocytes from hemolysis, secretion in glandular cells, and in the kidney. Human erythrocytes build up and maintain concentration gradients of 1:15 (between cell and serum). The corresponding value for potassium is 30:1. The numbers relate to the volume claimed by water. The membrane permeability for Na⁺ is smaller than for K⁺, which is perhaps due to the larger ionic radius of Na⁺ when including the water of hydration. 

Thanks to the numerous studies in this area, the processes surrounding the active transport of Na⁺ have already been largely elucidated. One can find an overview in SCHONER(32). The effect of fluoride on the Na⁺ transport has already been studied as well. OPIT(22) found about a 50% inhibition of the (Na⁺-K⁺) activated ATPase, when he added 4 x 10^-3M NaF. LEPKE and PASSOW(23) carried out measurement on so called "erythrocyte ghosts" and found that a strong K⁺ efflux elicited by F^- is accompanied by only a minimal Na⁺ influx. Unfortunately the F^- concentrations used by the authors were, at 4 x 10^-2M, much too high to be able to draw conclusions about the effects of physiological F^- concentrations. We therefore carried out experiments at significantly smaller F concentrations. To be precise, we studied the effect of different F^- concentrations on the Na⁺ exchange at the cell membrane after different times. We prepared the specimens for this experiment in a manner analogous to the calcium studies.  

It became evident that the relations were completely different from those for K⁺ and Ca²⁺ permeability. Figure 35 plots the dependence of the Na⁺ influx as a function of the NaF concentration, after three different times. The x-axis is measured in logarithmic units for the sake of a better overview. 

Figure 35 - ²⁴Na Content of Erythrocyte as a Function of the NaF Concentration 

After 1) 2 min. ; 2) 4 min ; 3) 15 min. 

Those cells that were exposed to a NaF concentration of 10^-4M contain the most Na-24 after 2 minutes. Saturation seems to be reached above 10^-3M NaF. The ²⁴Na concentration is smaller after 4 minutes, which means that a retrograde transport of the ²⁴Na that penetrated the cell must have started. After 15 minutes the maximum permeability has shifted to 5 x 10^-4M NaF. In addition, the intra-cellular ²⁴Na concentration has increased with respect to the shorter times. To elucidate these processes further, we depicted the course of the ²⁴Na concentration in the cell as a function of the time, with varying NaF concentrations. 

Figure 36 - Rate of ²⁴Na Erythrocyte Labeling vs NaF Concentration 

a) 0 M; b) 5 x 10^-5 M; c) 10^-4 M; d) 5 x 10^-4 M; e) 10^-3 M 

When no fluoride is added, the ²⁴Na concentration in the cell initially rises slowly until, after 10 minutes, it flattens out at saturation. The course is about the same at 5 x 10^-5M NaF, however a start up time of 6 minutes is needed, after which saturation quickly follows. A very strong influx of ²⁴Na into the cell occurs at 10^-4M NaF. The influx is apparently very rapid, since a retrograde transport had already occurred after the shortest time that we could record (2 min.). 

The course of this curve has a certain similarity to the movement of Na⁺ at nerve cells after a period of stimulation, although the processes take place significantly faster there. Unfortunately,  at this point we cannot draw any conclusions about the rate of the Na⁺ influx. In the case of the nerve cells, a strong Na⁺ influx is induced by the depolarizing effect of the ACh. The influx stops after a very short time (1 msec.), after which the Na⁺ that has penetrated is pumped back out with the help of a Na⁺ pump. When applied to this situation, this would mean that the Na⁺ permeability temporarily greatly increases due to the effect of the F^-, which was simultaneously added with the ²⁴Na. The cell would next have to compensate for this influx with some counter measure and following that pump the Na⁺ that has penetrated in back out. Oddly this "Na⁺-kick" decreases with rising F^- concentration. 

Of course the conclusions drawn here are highly hypothetical. Further and more differentiated recordings in this area might yield more information. It is, however, interesting that F^- concentrations of 10^-4M, and even lower, can exert a clear influence, which because of the central importance of Na⁺ permeability could possibly also be brought into connection with the observed vagotonic effects of F.  

d. Fluoride Permeability  

Of great importance for the permeability experiments is the question, at what rate and to what extent does fluoride penetrate into the cells? We therefore determined the distribution gradient of fluoride between the cell and the extra-cellular fluid with the help of the ¹⁸F tracer method. The absolute F concentration was 10^-4M. We also determined if an addition of ATP to the Serum (10^-3M) has an effect on the ¹⁸F distribution in a parallel experiment.   

This time we used a somewhat altered procedure. We again withdrew fresh blood from our own arm vein, whereby we again added 10% isotonic citrate solution. The ¹⁸F stock solution in twice distilled water was then brought to isotonicity with the serum by adding the corresponding amount of NaCl + NaF. The NaF concentration was thereby controlled in such a way that an end concentration of 10^-4M resulted after addition of the labeled solution to the blood (in the ratio 1:10). The erythrocyte volume of these specimens was now 38%. The two test tubes were placed in a thermoblock at 37⁰C. We started the stopwatch after addition of the labeled solution and, with the help of a glass capillary (Æ = 1mm), withdrew a blood specimen of 20 µl at different times. We quickly melted off one end of the capillary over the pilot flame of a Bunsen burner and centrifuged the specimen at 12,000 G with the help of a centrifuge from the "Eppendorf-Microliter" system. Maximal separation possible between the cells and the serum was thereby achieved within a matter of seconds.   

The time was recorded at the beginning of centrifugation. With the help of a glass cutter we now directly separated the erythrocytes and serum at the interfacial boundary and transferred the tubules into separate measuring vessels made of polyethylene with a length of 5 cm and a width of 2 mm. We stuck these through the lid of a test flask filled with Bray's solution, so that the radioactive specimen was completely below the surface of the liquid. This was possible because the g radiation of the ¹⁸F that arises from the annihilation of the positrons penetrated the wall on the inside of the vessel without difficulty and generated photons by way of mutual exchange with the Bray's solution. The photons were counted by the instrument. The yield was, however, greatly reduced with respect to the ß^- radiation. The advantage of this method was that the test flasks filled with Bray's solution could always be reused by replacing the inner plastic tubule. In addition, the need for elaborate specimen preparation was avoided. We then determined the distribution gradient a from the ratio of the volumetric share of the erythrocytes and of the serum, which in each case was derived by measuring the appropriate column length as well as the ratio of the radioactivity in the two phases. 

The results of the measurements are reproduced in the following table. 

Table 4. Erythrocyte/Serum Distribution of ¹⁸F vs Reaction Time 

The ¹⁸F distribution between the serum and the cell was not dependent on time, which means that the equilibration must have occurred before we started to record elapsed time. The F concentration in the serum remains twice as large as in the erythrocytes. Adding ATP to the serum raises the intra-cellular fluoride concentration. We can not yet cite a reason for this. Perhaps, however, the Ca²⁺ complexing characteristics of the ATP play a roll. The observation of the rapid exchange of fluoride at the erythrocyte membrane fits well with the observations of sodium exchange, which was influenced by the fluoride within very short times. 

e. Phosphate Uptake 

The conditions are different for phosphate ion uptake than for the ions considered so far, since the overwhelming proportion of the intra-cellular phosphate is present in organically bound form. At the pH of blood (7.4) about 75% of the phosphate is present as HPO₄^2- and about 25% as H₂PO₄^-. In addition to the procedural technique used so far to determine the distribution gradient of the radioactive substances between the serum and the erythrocytes, we used a new technique that allowed us to separate and identify the different labeled phosphate compounds in the erythrocytes in one procedure. We will dispense with a detailed representation of these studies at this point in order not to breakup the framework of this study. Instead, we are giving an overview of the most important results.  

First we added ³²P phosphate to fresh blood with added citrate and after a certain time separated the erythrocytes from the plasma. After hemolysis (dipping in liquid oxygen) we subjected the solution to high tension electrophoresis. We used a paper strip of 1 m in length and 15 cm in width soaked in a Veronal/HCl buffer of pH 8.6 for this procedure. The field strength was 40 V/cm. After one hour we developed the electrophorogram with the help of our radiochromatogram scanner. We could thereby separate eight radioactive phosphate compounds and identify them with the help of reference preparations, which we produced in radioactively labeled form through directed enzymatic conversion. 

If one labels blood by adding ³²P phosphate, inorganic phosphate can hardly be found in the cell, even after a short time. Instead, one finds ³²P ATP, ³²P-2, 3-Diphosphoglycerate (abbreviated 2.3-DPG), ³²P fructose-1, 6-diphospate (abbreviated FDP), as well as at least four additional glycolysis intermediates, however in smaller concentrations. The uptake of phosphate by the erythrocytes is therefore closely coupled with the glycolysis, just as the glucose uptake is. The radioactive phosphate distributes itself over all phosphate compounds, which are in equilibrium with each other. Once saturation is complete, the radioactivity of the individual substances is proportional to their absolute concentration. One can observe the dynamic of the incorporation of P-32 phosphate into the compounds in reference by tracking the labeling of the individual substances as a function of time.  

In order to study the rate of phosphate exchange between the plasma and the cell we next tracked the distribution gradient over function of time, whereby we simultaneously studied the influence of different fluoride concentrations. We determined the distribution gradient a according to the method indicated for ¹⁸F. We measured the radioactivity after precipitating the protein with trichloroacetic acid in aqueous solution using the "Cerenkov radiation" of the ³²P. 

Figure 37 - Distribution Gradient of ³²P Phosphate Between the Erythrocytes and the Plasma as a Function of Time 

1). No added fluoride; 2). with 5 x 10^-5M NaF; 3). with 2.5 x 10^-4M NaF 

The figure shows the temporal plot of phosphate uptake. Curve 1 describes the course as HEVESY(34) has already reported. Our measurements thereby agree quite well with his findings. Curves 2 and 3 show the same course in the presence of fluoride in concentrations that, at least in the case of curve 2, can still be described as physiological. One can recognize that the accumulation of ³²P in the cell, after a short start up period, rises almost linearly with time. After two hours the intra-cellular concentration is about the same as the extra-cellular concentration, however saturation is by no means reached yet. The concentrations of the phosphate compounds, which are in equilibrium with inorganic phosphate, must therefore be greater in the cell than in the plasma. Adding fluoride increases the phosphate uptake of the cell. After 7.5 minutes, however, this curve approaches the curve without added fluoride and then a stronger rise begins again. Based on the shape of this curve, the effect of the fluoride on the phosphate uptake by the erythrocytes seems to occur in two steps. At the same time one recognizes that even small fluoride concentrations elicit an influence, so that such an effect is possibly also present in physiological conditions. 

Next we studied the influence of fluoride on the established equilibrium during phosphate saturation. To do this we let erythrocytes in Ringer's solution come into contact with ³²P phosphate for 18 hours. After this time we separated the cells out and determined their radioactivity after hemolysis and precipitation of the protein with trichloroacetic acid.  

Figure 38 -  Erythrocyte Uptake of ³²P vs NaF Concentration After 18 Hours 

The phosphate uptake passes through a maximum at 2.5 x 10^-3M NaF. Otherwise, there are two opposing tendencies. At fluoride concentrations below 2.5 x 10^-3M, the effects that elicit an accumulation of phosphate in the cell dominate. At lower concentrations those effects that counteract this accumulation are dominant. A comparison with our glycolysis experiments shows that phosphate labeling decreases with the decrease in glycolysis activity of the cell. The formation of marked ATP also diminishes above 2.5 x 10^-3M. Perhaps the course of the curve means that at small F^- concentrations only those reactions that lead to a breakdown of organic phosphate are initially inhibited, while the formative reactions are hardly influenced yet. Upon further increase of the fluoride concentration these reactions are eventually also inhibited. Since inorganic phosphate is not stored in the cell, the distribution gradient can in this case take on at most a value of a = 1. With the help of the electrophoretic separation technique we could show that the concentration of labeled phosphoglycerate passes through the same maximum. The enzyme that further degrades these compounds is enolase, whose inhibition by fluoride WARBURG(21) has already described. Existence of the Mg fluorophosphate complex that he postulated could, however, until now not be directly proven.  

Upon examination of the glycolysis rate as a function of the fluoride concentration we found a stimulation of glycolysis at 2.5 x 10^-4 M NaF. Inhibition began only above this concentration and then increased continuously with rising F concentrations.

Introduction | Contents | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9