The Mobilization of the Calcium Carbonate Deposits in the Endolymphatic Sacs of Metamorphosing Frogs

The metamorphosis of amphibian tadpoles is a gradual process, but in the Anura there is a period called ‘metamorphic climax’ which is characterized by drastic morphological and physiological changes (Etkin, 1964). Before this metamorphic climax the animal is a typical tadpole with a beak-like mouth, a long tail and no external forelimbs. At the end of this period the animal resembles a small adult with a wide mouth, no tail and four well-developed legs. This re-organization of the meta-morphosing tadpole affects most organ systems and includes changes in the alimentary canal with the result that no food is eaten during the greater part of the climax.

The period of metamorphic climax is also a time when many of the bones of the frog start to ossify. Thus the skeleton is mineralized while the animal is isolated from dietary sources of calcium so that the calcium for this process must be obtained either by absorption through the skin or from some internal store. There is no satisfactory evidence available to suggest that either the tadpole or the adult frog can absorb calcium ions across the skin (Huf & Wills, 1951 ; Schlumberger & Burk, 1953; Krogh, 1938), but it has been suggested on a number of occasions that calcium carbonate may be resorbed from the endolymphatic sacs of tadpoles for this process (Guardabassi, 1952, 1953, 1956; Kreiner, 1954).

The endolymphatic sac is a diverticulum which arises, as the endolymphatic duct, from between the sacculus and utriculus of the inner ear with which it is in continuity. Typically the sac penetrates the wall of the otic capsule and comes to lie in the cranial cavity. In most vertebrates it is a small blind-ending vesicle which may contain a few crystals of calcium carbonate but which seems to be a relatively unimportant structure. In the amphibia and a few other vertebrates the endolymphatic sac expands within the cranial cavity to envelop portions of the brain. It reaches its greatest development in the Anura where it not only enlarges to form processes around the brain but also extends caudally along the vertebral canal and protrudes between the vertebrae of the adult. All parts of this system contain tiny crystals of calcium carbonate in the form of aragonite (Carlstrom, 1963).

The anatomy and development of the endolymphatic sacs of the common frog (Rana temporaria) have been described in detail by Whiteside (1922) who pointed out that these organs are very well developed and contain much mineral matter through-out most of the life of the tadpole. The suggestion that these deposits might form a store of calcium which can be used during metamorphic climax to form the skeleton is based upon an apparent reduction in the amount of calcium carbonate as assessed visually in cleared specimens or in X-radiographs (Guardabassi (1952, 1953, 1956, 1957), using Bufo bufo, B. vulgaris, R. dalmatina and R. esculenta-, Kreiner (1954) using Xenopus laevis.) With these techniques it is possible to show a large decrease in the contents of the endolymphatic sacs when the tadpoles undergo metamorphosis in a calcium-free medium (Guardabassi, 1957).

When Ca⁴⁵ is added to the water in which the tadpoles are reared it accumulates in the endolymphatic sacs of these animals. These specimens can then be transferred to running tap water when it is possible to demonstrate the presence of the radio-isotope in the bones of these frogs after metamorphosis (Guardabassi, 1959, 1960). It is not possible, however, to determine whether this result implies that the endolymphatic deposits are resorbed during metamorphosis or whether it simply demonstrates an exchange of calcium ions between different physiological compartments. On the basis of these observations it was decided to perform experiments which would attempt to provide information pertinent to the following questions :

1. Are the calcareous deposits in the endolymphatic sacs normally mobilized at metamorphic climax or does this only occur in tadpoles reared in water with a low calcium content?

2. Does the time at which this mobilization occurs coincide with the formation of bone mineral?

3. Is there a quantitative relationship between the amount of calcium resorbed from the sacs and that deposited as bone?

4. Is the low pH of distilled water a factor influencing the loss of mineral from the endolymphatic sacs, i.e. are the deposits involved in acid-base regulation?

The tadpoles of the common frog (Rana temporaria) were used throughout this work. They were all hatched from a single batch of spawn which was kept in tanks containing about 4 in. of aerated tap-water. After hatching the tadpoles were separated into eight batches of fifty-five animals and each batch was placed in 2 1. of tap-water in a plastic bowl. The temperature was kept at about 20° C. and the animals were fed on canned chopped spinach. Water and food were changed daily.

Although all the animals were of the same age there was some variation in their rate of development and it was consequently necessary to devise a comparative scale of development. This was done in two ways. First, use was made of the morphological stages described by Taylor & Kollross (1946) for Rana pipiens. Most of the characteristics described by these authors were found to be directly applicable to R. temporaria, although the duration of the stages differed somewhat. The second method was based on the use of the ratios of hind-leg : body length and tail : body length as an indication of development (Etkin, 1964). These ratios were determined from the specimens which were taken daily for analysis. The average value for each ratio on each day was then calculated. The changes in the ratios were plotted against time and the abscissa of this graph was then used as a relative scale of developmental days (Text-fig. 1). The important stage in our experiments is when the tadpoles enter the metamorphic climax. This is defined as the day when the first forelimb is protruded (stage XX in the Taylor & Kollross scheme) and is designated as day o on the relative scale.

Once Text-fig. 1 was constructed it was possible to give the tadpoles a ‘developmental age’ rather than a pure chronological age. The ‘developmental age’ was determined from both the Taylor and Kollross stage and from the ratios of hind leg : body lengths and tail:body lengths. By using the concept of ‘developmental age’ it was possible to remove some of the variation introduced into these experiments by the fact that the tadpoles developed at slightly different rates.

When the tadpoles had developed as far as prometamorphosis (Text-fig. 1) they were removed from the tap-water and placed in one of the four experimental media defined in Table 1. They continued to be fed on canned spinach and the water was changed each day. Random samples of tadpoles were taken from each treatment every 24 hr. and were preserved in 70% alcohol for later analysis.

The first step in the analysis of the tadpoles consisted of eviscerating the animal and dissolving the soft parts by repeated washings with warm 20% ethylene diamine. This was done in a 15 ml. centrifuge tube so that the remaining mineral deposits could easily be washed and separated. The inorganic matter was analysed for carbonate by means of a microdiffusion method similar to that of Conway (1962). The centrifuge tube containing the sample acted as the outer chamber of the microdiffusion unit and a loose-fitting glass tube with holes blown along one side was used as the inner chamber (Text-fig. 2). A solution of barium hydroxide (0-4 ml. O-IN containing 20% alcohol and thymol blue) was used as the carbon dioxide absorbent in the inner chamber and o·6 ml. N hydrochloric acid was added to the sample in the outer chamber immediately before the units were sealed. They were rocked for 2 hr. while the carbon dioxide liberated from the carbonate in the sample diffused into the inner chamber. The quantity of carbon dioxide liberated was determined by back titration of the barium hydroxide using O-IN hydrochloric acid from a micrometer syringe.

The dissolved sample in the centrifuge tube was washed into a crucible, evaporated to dryness and ashed at 400° C. to remove the last traces of organic matter. Ashing was necessary at this stage since the phosphate analysis which was to be used is very sensitive to traces of organic matter. The ash was dissolved in a minimum of hydrochloric acid and made up to 25 ml. with de-ionized water. A 10 ml. aliquot of this sample was analysed for phosphate by a colorimetric method using ammonium vanadomolybdate (Hansen, 1950; Kitson & Mellon, 1944) and a Spekker colorimeter with a 601 Ilford filter.

The phosphate was removed from a further 10 ml. aliquot of the sample by passing it through a short anion exchange column (Amberlite IRA 400) similar to that described by Lapidus & Mellon (1958). The effluent was collected and the column was washed with de-ionized water to give a 50 ml. sample. This was analysed for calcium using 15 ml. aliquots buffered at pH 12 and containing murexide as an indicator. The solution was ;titrated with 0·01 M ethylene diamine tetra-acetic acid (West & Sykes, 1960) and the end-point was determined photoelectrically using a 606 Ilford filter.

The total quantity of mineral in the tadpoles used in these experiments was in the range of 0·5−3·0 mg. and this had to be analysed for three different ions. Experiments were therefore conducted to determine the accuracy of the experimental methods in analysing samples of this size. The results are shown in Table 2. Many biological specimens contain contaminants which interfere with analytical techniques and to ensure that the methods used overcome these difficulties a comparison was made between the results of a microanalysis of a standard bone sample and a macro analysis of the same material (Table 3).

Once the mineral deposits of the tadpoles had been analysed for calcium, carbonate and phosphate ions it was necessary to relate these analyses to the size of the skeletal and endolymphatic stores. As a first approximation it is possible to consider that the bone is responsible for all the phosphate ions and the endolymphatic sac all the carbonate anions. It is known, however, that bone mineral contains a small amount of carbonate ions, and that the amount of this material depends upon the composition of the blood (Hirschman & Sobel, 1965). Samples of tadpoles’ bones were therefore analysed for calcium, carbonate and phosphate. The bones were obtained by removing the hind limbs of tadpoles of various ages and subjecting them to the treatments already described. The results are shown in Table 4 in the form of Ca :P, Ca : CO₃ and P : CO₃ ratios. Because the quantities analysed were very small these results should only be regarded as approximate and they are intended purely in order to provide a correction factor for the quantity of carbonate bound to bone. Using this information it is possible to estimate the quantity of calcium in the endolymphatic minerals by two methods. The following abbreviations are used :

The two estimates of calcium in the endolymphatic deposits agree well with a standard deviation of only 0·03 mg. Method A is, however, preferred as it depends solely upon the analyses of calcium and phosphorus which are more accurate than the analyses involving carbonate (Table 2).

The tadpoles used in these experiments were allowed to undergo metamorphosis in water of various pH and calcium levels. The results showed that pH had no significant effect upon the calcium deposits of the animals and the data have therefore been simplified to show only the variations between highand low-calcium treatments. The experimental period has been divided into ten 2-day intervals and the average values of endolymphatic calcium and bone calcium in each interval have been calculated and plotted in Text-fig. 3. Expressing the analyses in this way divides the results into three phases with turning points at day —4 (possibly — 6 in the low-calcium treatment) and day + 2. These phases have been used as the basis of a statistical treatment of the results in which the level of significance has been established by Student’s t test (Tables 5 and 6). In order to illustrate the discussion of the results the changes in the size of the various deposits have been plotted against developmental days in Textfig-4.

The X-radiographs of several tadpoles at various stages of metamorphosis are shown in Pl. 1. Three regions of the endolymphatic deposits are clearly shown, namely the otoliths and the cranial and spinal parts of the sacs. The results of many such radiographs have been used to assess, on an arbitrary scale of three units, the quantities of minerals in each of these regions. The results are shown by means of histograms in Text-fig. 5.

Two environmental influences were studied in this work. The acidity of the water in which the tadpoles underwent metamorphosis did not significantly affect the size of the mineral deposits in these animals. It appears therefore that the tadpoles are either isolated from the influences of the pH of the water in which they five or the calcium deposits are not involved in acid-base regulation to any detectable extent.

The concentration of calcium ions in the water greatly affects the amount of mineral deposited within the tadpoles even though all the animals were fed on identical diets of canned spinach. The tadpoles of R. temporaria are filter feeders (Savage, 1952) and it is concluded that considerable quantities of water are taken into the intestine with the food. Two solutions were used to study this calcium effect. The ‘low-calcium’ solution was actually prepared so as to be free of this ion but it seems likely that between the daily changes of the solutions traces of calcium may have entered the water from either the urine or the faecal strings of the tadpoles. The most obvious effect of the low-calcium solution was to lower the total calcium content of the animals to about 60% of that of the tadpoles reared under more normal circumstances. There was no increase in the calcium content of the tadpoles reared in either solution from day +2 until the end of metamorphosis (Text-fig. 4). This could be interpreted as being due either to a cessation of feeding or to the loss of a calcium pump from the skin or gills. The first possibility is preferred because of the lack of satisfactory evidence for an ion pump in the epidermis of tadpoles.

At the start of the experiment (i.e. day —8) there was no significant difference between the tadpoles in the two treatments (Table 5). By the end of phase 1 (Text-fig. 3), which occurs at day - 4 in the high-calcium regime and by day — 6 to - 4 in the low-calcium solution, the size of the endolymphatic deposits are significantly different. The tadpoles from the calcium-rich water contained considerably more mineral than the low-calcium animals, and this difference remained throughout the experiment whether the endolymphatic deposits were considered alone or whether ‘total hard parts’ were compared (Table 5).

The mineralization of the skeleton began to increase from day —4 onwards. A significant difference in the size of the treatments was not established, however, until after the end of the second phase (day + 2) when the tadpoles ceased to absorb additional calcium from the water.

These results show that the calcium available to the tadpole is stored in the endolymphatic sacs during premetamorphosis but that it ceases to be deposited there after day —4, when it is deposited directly in the bones. From days —4 to +2 there is no significant change in the size of the endolymphatic deposits (Table 6) indicating that this store of calcium carbonate is either not being resorbed during this period or alternatively it is being deposited at the same rate as it is being resorbed. The first possibility would again be the simpler explanation of the results.

The bones continue to be mineralized after day + 2 when the animals have ceased to feed. The rate of mineralization is greatest in those animals which had previously had access to most environmental calcium and bone formation continues throughout metamorphic climax in these specimens. The deposits of calcium carbonate in the endolymphatic sacs of these animals decrease by an amount which is slightly in excess of that required to provide all the calcium necessary for the continued formation of the skeleton. Thus both the temporal relationships and the quantitative analyses indicate that mineral is resorbed from the inner ear of the metamorphosing tadpoles to provide the calcium necessary for bone formation. There is a slight possibility that traces of calcium may continue to be absorbed from the intestine during this time since a small amount of food remains there even though the animals are no longer feeding. Analyses of the intestines of tadpoles at the start of metamorphic climax show, however, that they contain only one-tenth of the calcium necessary to account for the observed amount of bone ossification.

Those tadpoles which were raised in the low-calcium solutions ceased to form extra bone after day + 4. Despite this the endolymphatic deposits continued to be resorbed until the end of the experiment (Text-fig. 4). It appears therefore that from day +4 onwards these animals resorb calcium from their endolymphatic stores without depositing it in their skeleton. The calcium lost amounts to about 100-150μg. in 6 days. An almost identical quantity of calcium is lost from the endolymphatic sacs of the ‘high-calcium’ animals without being incorporated into the bones. This loss of calcium is clearly seen in the downwards slope of the total hard parts after day + 2 in Text-fig. 4. Since there are no faeces being produced from the alimentary tracts of these animals at this time it appears that this loss of about 20 μg. calcium/day probably occurs through the kidneys.

It should be realized that in these experiments the analyses of the calcareous deposits of the inner ear have included the otoliths as well as the material in the endolymphatic sacs. The otoliths, however, are involved in sensory functions and appear to be physiologically quite separate from the other calcareous deposits of the inner ear. Thus there is a small amount of calcium carbonate in the endolymph which is never resorbed during metamorphosis so that the amount of calcium remaining in the inner ear of the ‘low-calcium’ tadpoles probably represents the otoliths (Text-fig. 4). This implies that either one region of the membranous labyrinth is adapted to resist resorption (i.e. the otoliths) or another region is specialized to facilitate it (i.e. the endolymphatic sac. The evidence from the X-radiographs suggests that the second possibility is the more likely since there appears to be a gradient of increased resorption of the endolymphatic deposits from the posterior to the anterior of the sac (Text-fig. 5). In the mammals there is a high concentration of the enzyme carbonic anhydrase in the endolymphatic sac (Erulkar & Maren, 1960) and it would be of interest to know whether a similar situation exists in amphibians and whether the enzyme is involved in the resorption of the deposits.

The analyses of the tadpoles described in these experiments demonstrates that the resorption of the calcareous deposits of the endolymphatic sacs of R. temporaria is a normal event which enables the tadpoles to continue to ossify their skeletons even though they are unable to feed during the metamorphic climax. This phenomenon is presumably common to most anurans but it usually goes to completion, and is thus conspicuous only when the animals are allowed to metamorphose in distilled water (Guardabassi, 1957). In our experience the morphological changes under normal conditions are so small that the extent of the resorption is often concealed. This may have tended to obscure the conclusion that this resorption is probably a normal part of anuran metamorphosis.

One of us (J.B.P.) would like to thank the S.R.C. for a grant which enabled him to undertake this work.

PLATE I. X-radiographs of tadpoles at various stages of development. Animals 1-3 were reared in the ‘high-calcium’ solution and animals 4-6 are from the ‘low-calcium’ treatment.

In the early stages of metamorphosis there is an accumulation of mineral in the endolymph of both sets of animals and the otoliths (o.) and the endolymphatic sacs (e.s.) are clearly seen in animals 1 and 4. The deposits in the endolymphatic sacs of the ‘high-calcium’ animals remain clearly visible throughout metamorphosis (animals 2 and 3) but decrease rapidly in the ‘low-calcium’ specimens (animals 5 and 6). Note also that the bones are better ossified in animals 2 and 3 when compared with animals 5 and 6, but that the otoliths appear to be similar in both sets of animals and do not change during metamorphosis.

The assessment of the amount of mineral in the endolymphatic sacs of animals 3 and 6 is complicated by the ossification of the vertebrae. In the side views (3B and 6B) it can be seen, however, that the sac remains as a dorsal structure filled with mineral (arrowed) in animal 3 B, but that there is no corresponding deposit in animal 6B from the low calcium treatment.

* Basically no calcium apart from that derived from the animals’ excreta or their food.