Ionic Regulation in the Muscle Fibres of Carcinus Maenas: I. The Electrolyte Composition of Single Fibres

The electrolyte composition of animal tissues, particularly the excitable tissues like nerve and muscle, have interested biologists for many years, and many theories have been advanced to explain the fact that their composition is often considerably different from that of the surrounding blood.

For the most part attention has been devoted to vertebrate cells, but in recent years the most important advances in nerve physiology have come from a study of the axons of marine invertebrates. The giant axons of cephalopods, because of their very large size, have proved particularly favourable material (see review by Hodgkin, 1951).

For a study of muscle composition and ionic equilibrium the muscles of the Decapod Crustacea offer similar advantages since the individual fibres are much larger than those in vertebrate muscle and they can be easily separated from each other. The electrical activity of these muscles has already been investigated by Fatt & Katz (1953) who find important differences from vertebrate muscle.

For the work described in this series of papers the common shore crab, Carcinus maenas, has been used since, as it is a marine animal, the ionic concentrations are high and this makes for more accurate analysis and also the blood concentration and the equilibrium conditions can be altered by keeping the animals in diluted sea water. Muscle fibres derived from segments of the chela are used ; these fibres vary from 100 to 500 μ in diameter and are several millimetres in length.

The crabs were supplied by the Dove Marine Laboratory, Cullercoats, where they had been collected locally, and were transferred to an aquarium into which sea water was pumped from a closed circulation system. The temperature of the aquarium water was about 17°C. and the crabs would live in these conditions for many months if they were fed at regular intervals.

The composition of the aquarium water was slightly different from that of many natural sea waters. The concentrations of some of the main inorganic constituents are shown in Table 1, and a comparison of these figures with those for other sea waters (see, for example, Prosser, 1950, p. 78, table 9) shows the potassium and calcium concentrations to be unusually high. As will be seen later, these differences may possibly affect the blood composition of the crabs.

Blood was removed from the crabs in a pipette inserted through the arthrodial membrane at the base of one of the walking legs ; it was allowed to coagulate and 0·2 ml. of the plasma was diluted to 25 ml. and this solution was used for the measurement of blood conductivity, sodium, potassium and chloride concentrations. A second 0·2 ml. blood sample was taken for the estimation of calcium and magnesium concentrations.

Sodium and potassium were measured in solutions diluted to contain not more than 4 and 8 p.p.m. respectively, by means of an ‘Eel ‘Flame Photometer. Chloride concentrations were measured by the Volhard back-titration method using N/100 sodium thiocyanate, and this titration method was also used for the estimation of calcium and magnesium after they had been converted quantitatively to chlorides by means of the method described by Shaw (1955). If sufficient calcium was available, this ion was also measured by the Flame Photometer in a solution containing up to 40 p.p.m. of the ion.

For this investigation the two muscles in the meropodite segment of the chela were used. These are the carpopodite extensor and flexor, and they lie more or less parallel to each other along the length of the segment. The fibres of these muscles are inserted on the inside of the shell of the meropodite and have their origin on the chitinous tendon which is attached to the junction of arthrodial membrane and the carpopodite. The extensor was used most often as the fibres are larger and longer than those of the flexor. The muscles were prepared by bisecting the meropodite longitudinally so that each part contained one of the muscles; the muscle was then removed by cutting the insertions of the fibres on the shell with a fine scalpel and finally severing the top of the tendon so that the entire muscle, together with the tendon, could be removed. The excised muscle was blotted on filter-paper to remove as much as possible of the adhering blood and then washed quickly in a solution of dextrose isotonic with the blood to remove any remaining blood. For this operation the tendon was held in forceps and the fibres agitated in the solution with a blunt needle rapidly so that the total washing time did not exceed about 30 sec. After this washing the muscle was again blotted on filter-paper to remove the dextrose solution. Now the tendon was removed by severing, with a sharp scalpel, the attachment of the muscle fibres. The muscle prepared in this way was used for the estimation of ion concentrations in the whole muscle.

The individual fibres of the muscle are large and well separated from each other so that single fibres can be easily dissected. Muscles dissected from the meropodite as described above and washed in isotonic dextrose were transferred to liquid paraffin. The cut end of a single fibre was held in forceps, the fibre separated from the rest of the muscle and its attachment to tendon cut with fine scissors. The fibre was lifted out and blotted gently to remove liquid paraffin.

For the measurement of the electrolyte composition of the whole muscle, the prepared muscle was first weighed and then dried to constant weight to determine the water content. For cation estimations the dried muscle was incinerated in a muffle furnace at 450−500°C. until all of the organic matter had been removed. The ash was dissolved in N/10-HCI and sodium, potassium, calcium and magnesium concentrations measured in the same way as described for the blood.

Muscle chloride was measured by the Volhard method. The dried muscle was placed in a crucible, together with a known quantity of silver nitrate and about the same volume of concentrated nitric acid, and this mixture heated in a 100°C. oven until the muscle had dissolved. The presence of the silver nitrate prevented the loss of chloride which would have otherwise occurred.

Single muscle fibres, isolated for analysis as described above, were each weighed on a 5 mg. torsion balance, the average weight being about 0·5 mg. For chloride, calcium and magnesium measurements the weighed fibres were transferred to small ‘Hysil ‘tubes, 2 cm. long and 2 mm. internal diameter. Concentrations were estimated using the ultra-micro methods described by Shaw (1955) which involves the quantitative conversion to chlorides and the estimation of the chloride by the Volhard method. Potassium estimations were made in the same way except that the incineration was carried out in very small platinum dishes and the ash subsequently washed into the ‘Hysil’ tubes with dilute HC1. The analyses were arranged so that the final titration measured between 0·5 and 1 μg. of chloride and the accuracy of the methods about ±1%. Fibre chloride was titrated in the same way after digestion of the fibre in nitric acid in the presence of excess silver nitrate.

Conductivity of the blood was measured in a conventional conductivity cell which contained 5 ml. of fluid. The cell formed one arm of an impedance bridge supplied with an alternating voltage of 1500 eye./sec. The resistance of the cell was measured when filled with a blood sample diluted to contain about 20 p.p.m. of salt.

An estimate of the conductivity of single muscle fibres was made by allowing the fibre to act as the conductivity cell itself. The electrodes were formed by small strips of platinum foil fixed on to a block of wax as shown in Fig. 1. Wires attached to the base of these strips were connected to the bridge arm. A single fibre was laid across the two electrodes and a large drop of crab Ringer solution placed on the strips at each end of the fibres to make a low resistance contact with the electrodes. The specific resistance of the fibre was calculated from measurements of the resistance of the preparation and the dimensions of the fibre (length and diameter) made to the nearest $11 0 0$ mm. by means of a travelling microscope. The method was checked by using, in the place of the fibres, pieces of glass capillary of roughly the same dimensions and containing saline solutions of known conductivity. As for the fibres, the resistance of the fluid and the length and diameter of the capillary were measured. The calculated specific resistances for two solutions are shown in Table 2.

The mean values for the two solutions of 20·6 and 40·9 agree quite well with the calculated values of 20·3 and 39·0 respectively, but the accuracy is not of a very high order. Standard deviations of ± 3·3 for the Ringer solution and ± 4·9 for the 50% Ringer solution are found and thus no great reliance can be placed on a single determination made by this method, but six or more measurements will reduce the error to acceptable proportions.

Measurements of the muscle fibre membrane potentials were made by means of intracellular electrodes of the type first used by Ling & Gerard (1949). These were constructed from 2 mm. ‘Pyrex’ tubing, one end of which was drawn out to a fine capillary by hand and a very fine tip being drawn on the end of this by the use of a De Fonburne Microforge. The diameter of the tip was 2μ or less. The electrode was filled with saturated KCl solution and a silver wire coated with silver chloride inserted through the other end of the electrode, the completed electrodes having a resistance between 2 and 10 MΩ. Then the electrode was held in a micromanipulator of the Chambers type and was used to pierce the fibre membrane ; the other electrode which remained on the outside of the fibre was constructed in the same way except that the capillary at the tip was of much greater diameter and hence the resistance of the whole electrode much lower.

The fibre potentials were measured on intact carpopodite extensor muscles. The carpopodite and the meropodite were removed from one of the chelae and the shell of the meropodite cut longitudinally so that the flexor was removed leaving the extensor intact (Fatt & Katz, 1953). The preparation was fixed into a ‘Perspex’ box, just large enough to accommodate it, by means of small pieces of plasticine. The external electrode was held by one of these pieces so that its tip was near the muscle. The preparations was covered with blood which had been extracted from the animal previously, allowed to coagulate and the precipitate removed, and finally a thin layer of liquid paraffin was spread over the blood. The box was placed on the stage of a Greenough binocular microscope and illuminated from below. The intracellular electrode was lowered beneath the paraffin layer and watched while it was inserted transversely through the fibre membranes.

The potential difference between the two electrodes was measured by a valve millivoltmeter consisting of a pair of RCA 954 valves arranged as electrometers in a bridge circuit and the readings were made on a C.I.C. spot galvanometer connected across the two anodes. The input impedance of the apparatus was 1000 MΩ. The calibration of the apparatus and the electrodes was checked before and after each set of readings by placing the two electrodes in separate vessels each containing crab Ringer solution and then applying a variable known potential between the two vessels.

The ionic composition of the blood of Carcinus maenas has been thoroughly investigated by Webb (1940). However, in view of the importance of an exact knowledge of the blood composition in relating it to the concentrations of ions found in the muscles, measurements have also been made on the blood of the crabs used in this work. Table 3 gives the summarized results of these analyses and indicates certain differences from the results obtained by Webb. The chloride concentration is slightly lower and the sodium concentration distinctly less, but on the other hand magnesium is higher and calcium very much higher (40 %). Some of these differences may be due to the unusual composition of the aquarium water—for example, the high blood calcium might be related to the high calcium concentration of the latter—but it seems unlikely that the aquarium water composition would affect the concentration of sodium and chloride in the blood. The total cation concentration agrees well with the conductivity measurements which are considerably lower than those for the aquarium water (552 compared with 610). These differences are probably due to other factors such as the different localities from which the crabs were collected, their size and age, etc.

Measurements have been made of the potassium, chloride, calcium and magnesium concentrations in single fibres. The fibres were isolated from the muscle as described above, and then the remainder of the muscle was dried and analysed itself, so that the concentration of ions measured in the single fibres and in the prepared whole muscle could be compared. The water content of each fibre was not measured but was calculated from the wet weight of the fibre and from the water content of the whole muscle as determined from the wet and dry weights of the remainder of the muscle. This averaged 73·5% of the wet weight.

The analyses of the chloride content of thirty muscle fibres derived from the muscle from five crabs are shown in Table 4. The mean value is 53 mM./kg. of fibre water; there is little variation in the mean values for the individual crabs and the differences are not statistically significant. The variation in the chloride concentration in the individual fibres of a single crab is more marked, and since an average figure was used for the water content this may indicate either that the fibres vary in their water content or in their chloride concentration. Probably the results reflect a variability of both factors.

Table 4 also shows that there is a close correspondence between the chloride concentration as determined in single fibres and in the whole muscle, and thus indicates that the method employed for the preparation of the muscle for analysis leaves it practically free from contamination with extracellular material such as blood.

The value for the chloride content is much higher than is found in vertebrate muscle—for example, Boyle & Conway (1941) give the chloride concentration of frog’s sartorius muscle as 1·2 mM./kg. H₂O—but this is to be expected in a marine organism. It may be compared with the chloride concentration found in the giant axons of cephalopods, e.g. Loligo axon 40 mM./kg. H₂O (see Steinbach, 1941a; Hodgkin, 1951, table 2).

The concentrations of potassium in the fibres, shown in Table 5, were measured on twenty-nine isolated fibres from five crabs. Again the mean concentration for individual crabs are not significantly different but the potassium concentration of the individual fibres exhibits considerable variation. The mean value of 112 mM./kg. fibre water again is not significantly different from the value found in the analysis of the whole muscle.

It is noteworthy that although the chloride concentration is much greater than that of vertebrate muscle, the potassium concentration is of the same order. Thus Boyle & Conway (1941) give the concentrations of potassium in frog’s Sartorius as 125 mM./kg. H₂O, and similar figures have been found for mammalian muscles.

Concentrations of potassium in muscles of other marine animals are not well established. Steinbach (1940, 1941 b) gives values for Thyone muscle and Phascolo- soma muscle as 169 and 106 mM./kg. tissue respectively, but these analyses include from 30 to 40 % extracellular material. If corrected for this and for the water content the values expressed in mM./kg. H₂0 would be considerably greater than found in Carcinus muscle. In the nerve fibres of several marine animals the analyses are more certain and all show very high potassiun concentrations—for example the giant axons of Loligo (Steinbach & Spiegelman, 1943 ; Keynes & Lewis, 1951 a), the axons of Sepia (Keynes & Lewis, 1951a) and the nerve fibres of Carcinus (Keynes & Lewis, 19516) have potassium concentrations between 330 and 460 mM./kg. water. It would appear that this difference in potassium concentration between that found by Keynes & Lewis for Carcinus nerve, and that reported in this paper for the muscle must be associated with some functional difference in the behaviour of the ions in the two different cell types.

The concentrations of calcium and magnesium were determined together on the isolated fibres. Owing to the low concentration of calcium present, in most cases several fibres (generally about six) were isolated, and used together for each determination. This can be seen in the weight records which are in general much greater than for single fibres. Table 6 shows the results for twenty-five groups of fibres derived from five crabs. The mean value for calcium is 5·2 mM./kg. water and for magnesium 16·9. These values do not differ much from the concentrations found by Boyle & Conway for the frog’s sartorius muscle 3·3 and 16·7, although in Carcinus these concentrations are lower than the respective concentrations in the blood, while the opposite is the case in the frog.

It has been seen in the previous paragraph that the results of the single fibre analysis are not significantly different from those found for the whole muscle which had been prepared by careful washing. Although losses of ions from the muscle would almost certainly occur in isotonic sugar solutions (e.g. Steinbach, 1941a) the washing was always rapid (less that 30 sec.), and it is not likely that much change would have occurred during this period. The summarized results for all the measurements made on the whole muscles are shown in Table 7 which includes estimations of sodium concentration, which were not made on the single fibres. However, in the latter case, in view of the close correspondence between the analysis for single fibres and the whole muscle for the other ions, there is little doubt that these analyses do represent more or less truly the concentration of sodium found in the muscle fibres. Table 7 shows the total cation concentration to be 224 m.equiv./kg. H₂O, and conductivity measurements (see below) indicate that there is no reason to suppose that any other cations exist in the fibre water. This figure is much less than half the corresponding figure for the blood, and if it can be assumed that the muscle fibres are in osmotic equilibrium with the blood then a large number of unionized molecules must also be present. This is not unusual among marine animals—a fact that was emphasized 50 years ago by Fredericq (1904). It is interesting in this connexion that recently Camien, Sarlet, Duchateau & Florkin (1951) found in the muscles of lobsters considerable quantities of free amino-acids of which proline, glycine and arginine were present in concentrations greater than r%.

The difference between the total cation concentration and the chloride concentration (which was the only anion measured) is very marked, the deficit in the anion fraction being 170 m.equiv./kg. Some of this will be accounted for by other inorganic anions such as bicarbonate, sulphate and phosphate, but the majority must be as organic anions. Whether this fraction consists of protein and organic phosphate compounds alone or whether other organic acids such as the acidic amino-acids are present remains to be discovered.

Conductivity of thirty isolated fibres from four different crabs was measured and the results are shown in Table 8. As pointed out earlier, no significance could be attached to individual results and these are therefore not given. The agreement in the results for the four crabs is good, but the variability in the apparent conductivity of the fibres of one crab is greater than found in the test experiments with glass capillaries. This may, to some extent, express true differences in the individual fibres, but may also be due to the difficulty of assessing accurately the true dimensions of the fibres due to changes in diameter along the length, which were often found.

To interpret the mean value for the fibre specific resistance exactly in terms of ionic concentrations is not possible until the nature of all the muscle ions and their ionic mobilities are known, but an approximation can be made.

Now the concentration will not be as low as 161 m.equiv./l. since ions of lower mobility than potassium (e.g. sodium ions) are present, but on the other hand is not likely to be greater than the concentration of the NaCl solution. Now the analyses given in the previous section showed that the total concentration of cations was 224 m.equiv./l. and thus the conductivity measurements indicate that more than 72% of these cations are present in an ionized form and possibly as much as 91 %. The divalent ions, calcium and magnesium, account for about 22% of the cations, and in view of known facts that these ions tend to form undissociated compounds with proteins and other large molecules it is possible that 50% or more of these molecules may not contribute to the conductance of the fibres.

In order to be able to interpret the differences in concentration of the ions in the muscle fibres and in the blood measurements have been made of the potential difference across the fibre membrane which separates the two. A comparison is then possible between this potential and the calculated equilibrium potentials. The membrane potentials of a very large number of fibres were measured on healthy and active crabs, and the results are summarized in Table 9. The consistency of the results is good, although lower potentials have been recorded in less active crabs.

The mean value of 58 mV. is somewhat lower than recorded by Fatt & Katz (1953) for the fibres of Portunus depurator, and occasionally Carcinus maenas, for which they found a mean value of 70 mV. This difference may be due to several factors, such as the slightly lower temperature, the different species used for most of the experiments and the different physiological conditions (Fatt & Katz used a crab Ringer solution rather than the animals’ own blood).

The potentials recorded in this paper for Carcinus muscle are much lower than those found in vertebrate muscle using the same micro-electrode technique (see Hodgkin, 1951, table 1), where potentials of around 80 or 90 mV. have been recorded but resembles more the potentials found in the cephalopod axons.

Table 10 shows the values of the equilibrium potentials calculated from the ratios of the ions found in blood and muscle and also the observed membrane potentials. The equilibrium potentials have been calculated for all the ions using the mean values from Tables 3 and 7. In addition, in some experiments membrane potentials and potassium and chloride concentrations were measured in the same crab, and these are also given in Table 10, together with the calculated equilibrium potentials.

In the case of potassium and chloride the correspondence between the equilibrium potentials and the measured membrane potential is very close both for the mean values and for the individual cases. Crab 31 should be noted since in this case the crab was not very active and the membrane potential was low, and it is seen that the equilibrium potentials are also correspondingly low. There is thus no necessity to postulate any activity on the part of the fibre or its membrane in maintaining the concentrations of these ions which are found in the fibre.

With the other cations the situation is different, for sodium, calcium and magnesium all require equilibrium potentials of the opposite direction from the observed membrane potentials. In the case of the divalent ions, since there is doubt if they are present in a completely ionized form and since nothing is known of the permeability of the membrane to these molecules, no deductions can be drawn other than that if any equilibrium does exist between those in the fibre and the blood it is certainly not a simple one.

With sodium the position is somewhat clearer since it has been invariably found that sodium ions are present in low concentrations in muscles and nerves of both vertebrate and invertebrate animals. In those cases where the penetration of the muscle by sodium has been measured using radio-active sodium as a tracer the fibre membrane has always been found to be permeable (e.g. frog’s sartorius; Levi & Ussing, 1948; Harris & Burn, 1949; rat gastrocnemius muscle, Heppel, 1939; Sepia axons, Keynes, 1951), and although measurements have not yet been made on Carcinus muscle it seems likely that the same will be found here. This means that the low concentration sodium ion in the muscle fibres must be maintained by some active process such as the sodium extrusion mechanism first suggested for vertebrate muscle by Dean (1941).

Boyle & Conway (1941) proposed, as a result of their experiments on the frog’s sartorius muscle, that the muscle fibre membrane was permeable to the smaller ions such as potassium and chloride but impermeable to larger ones like sodium and organic anions. They further suggested that the distribution of the potassium and chloride ions was the result of a Donnan equilibrium which existed across the fibre membrane as a result of the high concentration of sodium ions on the blood side of the membrane and the large number of organic anions in the fibre. The use of radioactive sodium has shown that the fibre membrane is in fact permeable to sodium ions, and in consequence this theory has had to be modified to take into account this fact. Dean (1941) proposed that sodium ions were actively extruded from the fibre and that this ‘sodium pump’ balanced the intake of sodium ions by diffusion.

The results obtained for the muscle fibres of Carcinus, as far as they go, are in complete agreement with the predictions of Conway’s theory, despite the fact that the ionic concentrations both in the blood and in the fibres are, in most cases, totally different from those found in vertebrate muscle. Whether or not this agreement is a real one and not merely coincidental will be shown by the composition of the muscles under conditions of varying blood concentration. It is under these conditions that it can be seen to what extent the fibre behaves in a passive manner or actively regulates the many variable factors, such as ion concentrations, water content and membrane potential.