1. Trypan blue, at concentrations around 0·025 per cent., was allowed to act on cleavage, gastrula, and early neurula stages of Xenopus, Axolotl, and Triturus alpestris and palmatus for various periods (about 24 or 48 hours).

  2. A variety of abnormalities were produced, of which the most striking were: (a) Oedema, (b) Microcephaly, (c) Prevention of elongation of the tail bud. (d) Vacuolation and swelling of the epidermal cells and fusion of the epidermis to the mesoderm, (e) Degeneration of the neural tissue (and to a lesser extent of the mesoderm) with appearance of vacuoles in the cells, (f) Suppression of gastrulation. (g) Mesodermalization of the notochord, similar to that produced by lithium.

  3. It seems unlikely that all these effects are consequences of one single original action. Some of the abnormalities (particularly (d) and (e), and to a lesser extent (b) and (g)) can be produced when the exposure to the dye occurs after gastrulation.

  4. In view of the lithium-like effect, Professor S. Ranzi of Milan has tested the action of trypan blue on the viscosity of protein solutions and their resistance to the action of urea, and found that it behaves similarly to other ‘vegetalizing’ agents.

  5. The incidence of these types of abnormality differs in the different amphibia investigated, as it also appears to do in the various stocks of mice which have been studied.

Considerable interest has been aroused by the report of Gillman, Gilbert, Gillman, & Spence (1948) that the vital dye trypan blue when injected into the pregnant rat brings about the appearance of various abnormalities in the offspring. The mammal embryo is notoriously difficult to attack experimentally, and trypan blue therefore seems to provide a valuable tool which, moreover, might throw light on the mechanisms of the important effects which are known in some cases to be produced on the foetus by pathological states in the mother. Further studies on the effects of injected trypan blue on mammalian embryos have therefore been made by Hamburgh (1952, 1954), Waddington & Carter (1952, 1953), and Murakami and his collaborators (1952, 1954), all of whom worked on the mouse, and by Harm (1954) who studied the rabbit. All these authors agree, on the whole, in the type of results which they have obtained, although there are some minor differences between their reports which will be considered later. All the investigations, however, have revealed a rather diversified spectrum of malformations from which it is by no means easy to deduce whether trypan blue has one primary effect (and if so, what is its nature) or whether it operates in a number of different and rather unspecific ways. Moreover, the method of administration employed in mammals, namely injection into the maternal blood-stream, leaves it open whether trypan blue can have a direct effect on the embryo or whether its action is always meditated through some alteration produced in substances supplied by the mother to the foetus.

In the hope of obtaining further information on these points trypan blue has been administered to amphibian embryos of a number of types.

The dye (taken from the same purified samples, supplied by Messrs. Imperial Chemical Industries, Dye Stuff Division Ltd., which were used in the investigation described by Waddington & Carter, 1953) was made into a 1 per cent, solution in distilled water and this solution further diluted with one-fifth Holtfreter solution. Embryos of Xenopus, Axolotl, and Triturus palmatus and alpestris were placed in solutions of appropriate dilution for various lengths of time before being transferred into one-tenth Holtfreter. The embryos were usually left in their vitelline membranes. A total of about 700 embryos were studied, about 80 being sectioned for microscopical examination.

1. Xenopus

Eggs and embryos of Xenopus, removed from the jelly but still within the vitelline membrane, were treated with solutions of the dye in one-fifth normal Holtfreter solution. After some preliminary experiments a dye concentration of 0·025 per cent, was found to be suitable. Eggs of various ages from the early cleavage to the neurula were left in this for periods of 24 or 48 hours, after which they were transferred to one-fifth normal Holtfreter solution for further cultivation. Embryos which remained healthy showed only slight signs of staining by the dye, but dead cells coloured intensively. In most batches of eggs subjected to the dye there was considerable variation in the degree of abnormality produced, probably as a result of differences in the degree to which the dye-stuff was taken up by the living material. On the whole the effects were stronger the earlier the stage at which the egg was placed in the dye, and the longer it remained there.

In some of the eggs treated in early stages the process of gastrulation was completely inhibited. The end result of development was the formation of a ball of rather featureless cells, which were heavily yolk-laden and more endodermlike towards the centre of the mass, and more epidermis-like towards the outer surface, which was often strongly wrinkled. In some regions the superficial layer of tissue was lifted away from the more deeply lying cells leaving a fluid-filled vesicle. In less extreme cases invagination took place at least to some extent. The resulting embryos often exhibited various degrees of reduction of the head, as well as irregularities in the tail, and were very often affected by a severe oedema. An example is shown in Fig. A of the Plate. On sectioning, such embryos showed a variety of abnormalities. Perhaps the most interesting of these consisted in the reduction of the notochord. In some instances (Plate, fig. B) this might be completely absent, the somite mesoderm from each side uniting in the midline to give a very irregularly segmented mass. In such cases the neural tube was also very much reduced, and in the example figured, which is from an embryo which showed very little axial elongation, the epidermis is thickened and wrinkled. In another example (Plate, fig. C), a notochord is present though it is much smaller than normal, and the somite mesoderm unites beneath it. The embryo is oedematous and the tissue of the neural tube is undergoing a peculiar type of degeneration in which many small vacuoles appear among the cells. Similar vacuoles appear in the mesoderm where, however, they are a normal feature of the histogenesis of the tissue.

The suppression of the notochord in embryos treated before gastrulation is probably brought about by an effect on the processes of regionalization by which the invaginated sheet of mesoderm becomes divided up into a strand of notochord in the middle with somite mesoderm on each side. This process is a ‘field phenomenon’, as Yamada (1940) in particular has shown, and the effect of trypan blue treatment of early embryos seems to be to reduce the ‘potential’ of the central strand of mesoderm so that it develops into somite instead of chorda. Even when embryos are not placed in the trypan blue until the end of gastrulation, a certain degree of mesodermalization is sometimes produced; an example of this is shown in Fig. D of the Plate. With still later treatments, the notochord differentiates normally.

The development of the neural system is influenced in several different ways. In the first place, when there is extensive suppression of the notochord the overlying neural system is also extremely reduced. This reduction can probably be regarded as a result of a weakening of the inductive powers of the mesoderm. A different type of action is probably responsible for the production of microcephaly, which seems in most cases to result from an inhibition of the forward movement of the invaginating mesoderm, accompanied by a depression of the anterior part of the inductive field. This is a very usual result of inhibitory agents acting during the process of gastrulation. Trypan blue, however, also exerts a third type of influence on the neural system, namely a toxic effect on the differentiating cells, which in some embryos are obviously unhealthy and show a large number of small non-staining vacuoles in the cytoplasm (Plate, fig. C).

2. Axolotl

A small series of axolotl eggs were exposed to 0·025 per cent, trypan blue for hours in blastula and early gastrula stages. One day after their removal from the dye solution some of the embryos were found to be completely exogastrulated. Even those which developed best exhibited extreme grades of microcephaly, no eyes at all being present (Plate, fig. E). At the posterior end several of the embryos had a large swelling, which sections showed to be composed mainly of notochord. These two effects are probably due to a partial suppression by the trypan blue of the gastrulation movements and of the elongation of the tail-bud. The elongation, particularly of the posterior end, was also inhibited in embryos treated from the late gastrula stage onwards. Sections from other individuals treated as blastulae or young gastrulae showed some clear examples of mesodermalization of the chorda. In extreme cases the chorda was completely absent (Plate, fig. F); in others, although the chorda was present, it was improperly differentiated, the cells remaining full of yolk granules to a much later stage than usual. There were also some morphological abnormalities in the anterior part of the nervous system (Plate, fig. H).

3. Triturus palmatus and alpestris

In a series of eggs of two species of Triturus, treated with similar concentrations and exposures, the effects were in general less striking, and more suggestive of a general toxicity of the substance than of any specific effects. In some eggs treated from early stages, gastrulation was completely suppressed, the embryos remaining more or less featureless balls of cells. Sometimes these showed a certain thickening of the outer layer along a line which presumably corresponds to the axis. There was little sign of any differentiation of specific types of tissue within this linear thickening. One can perhaps interpret these as embryos in which gastrulation had taken place but in which there was an extreme degree of mesodermalization, leading to the complete suppression of notochord formation and of the induction of the neural system. The embryos which were less severely affected were usually relatively normal, although considerably slowed up in their rate of differentiation. In a few cases there was a slight reduction in the size of the chorda, accompanied by a tendency for the somite mesoderm to unite beneath it; this represents a mild degree of mesodermalization. The embryos are frequently oedematous, but showed little tendency towards microcephaly. The epidermis is often highly abnormal, the cells becoming very large and vacuolated (Plate, fig. G). In the axial region the epidermis, mesoderm, and neural tissue are often not properly separated from one another and in such cases the neural tissue is usually badly differentiated. There were also some examples of disturbed morphogenesis of the neural system, particularly of the anterior region.

These experiments show that trypan blue can cause abnormalities in development when administered directly to embryos, and does not need to operate through some intermediary system such as the maternal blood-stream. It remains an open question whether when it is injected into pregnant mammals it is absorbed by the foetus and acts directly on it, or whether it operates by modifying the maternal fluids, but there is certainly no a priori reason for making the latter assumption. In mammalian embryos whose mothers have been injected, there is no visible sign of coloration by the dye except in the yolk sac and placenta, and this might suggest that the dye is not affecting the embryo directly. However, there is very little sign of any blue tint in amphibian cells exposed to trypan blue, at least until they become necrotic, when the dye is taken up quite rapidly; it seems, therefore, that the dye can produce a teratogenic effect when its intra-cellular concentration is extremely low.

In the experiments with amphibia a number of different kinds of malformations have been produced. The main types are:

1. Oedema

The treated embryos, particularly of Xenopus, are frequently extremely swollen and full of fluid in the tissue spaces.

2. Microcephaly

The main cause of this would seem to be an inhibition of the gastrulation movements. In embryos treated at the neural plate stage, however, in which the invagination would be complete, there are also signs of bad differentiation of the head, and it seems probable that the anterior part of the neural system is more sensitive than the posterior to some relatively unspecific inhibitory influence on tissue differentiation.

3. Formation of large masses of notochord at the posterior end

This effect has been seen both in Xenopus and Axolotl. It is presumably due to an inhibition of the elongation of the tail-bud.

4. Abnormalities of the epidermis

The cells of this tissue frequently become vacuolated and swollen. The epidermis may also become fused with the underlying mesoderm. It is sometimes thrown into numerous folds and wrinkles, probably as a result of continued growth in an embryo which is itself failing to become elongated.

5. Vacuolar degeneration of neural tissue and mesoderm

Various forms of tissue necrosis have been seen, of which this is the most striking. It is particularly marked in Xenopus, whereas in Triturus a more usual appearance is pycnosis of the nuclei.

6. Suppression of gastrulation

In some eggs treated before gastrulation and very severely affected, the entire process of invagination is suppressed and the embryo becomes a more or less featureless lump in which no tissue differentiation takes place. This can be regarded as an extreme form of the disturbances of gastrulation and morphogenetic movement which bring about the microcephaly and posterior enlargement of the chorda described in paragraphs 2 and 3 above.

7. Mesodermalization of the notochord

Both in Xenopus and Axolotl, but not in the comparatively small series of Triturus, examples have been seen of an effect extremely similar to the well-known action of lithium on the amphibian embryo. The notochord is partially or completely suppressed, the material from which it should develop being converted into more lateral mesoderm, of the nature of somite or even pronephros. In such embryos the two rows of somites are usually united across the midline, and the neural tube is also very much reduced, probably as a consequence of the reduction in the axial mesoderm. It is noteworthy that in Xenopus even with treatment as late as the end of the gastrula stage there may be some apparent mesodermalization of the notochord, but perhaps this should be compared rather with the ‘Chorda blocking’ described by Hadorn (1951) than with the normal lithium effects.

The mesodermalizing action of lithium is usually regarded as a relatively specific effect rather than a sign of a general inhibitory action. It has been compared with the vegetalization produced by the same substance on the echinoderm egg. The mechanism of these lithium effects has been considerably discussed (e.g. Lehmann, 1945; Pasteels, 1945; Gustafson, 1950, &c.). One of the most plausible hypotheses has been advanced by Ranzi (cf. 1951; Ranzi & Citterio, 1954). He has described a number of substances which act similarly in vegetalizing the sea-urchin embryo, while there is another group which shows the opposite action, increasing the tendencies towards animal development. It is known that at least one member of the former group, namely lithium itself, produces a mesodermalization of the chorda in amphibia, while some of the latter group (e.g. urea) tend to cause an increase in the size of the amphibian chorda.

Ranzi suggests that both groups of substances act by virtue of their influence on the state of aggregation of protein particles in solution. He finds that vegetalizing substances, at the concentrations at which they are embryologically active, bring about an increase in the viscosity of fibrillar protein solutions, and at the same time render the particles more resistant to the demolizing action of urea. The animalizing substances have the opposite effects. This hypothesis is based primarily on the properties of substances known to act in a vegetalizing manner on echinoderm eggs; rather few such substances have yet been shown to produce mesodermalization in the amphibia. It therefore seemed interesting to test the effect of trypan blue on protein solutions in the manner that Ranzi has previously used. Fortunately, Professor Ranzi himself expressed his willingness to carry out these tests, and we are extremely grateful to him for allowing us to quote the following results.

Euglobulin a + b from frogs’ eggs was prepared according to the method of Ranzi & Citterio (1954). Six samples each of 10 ml. were taken, and to each was added 10 ml. of 1 M KC1 containing various concentrations of trypan blue, so that the concentration of the dye in the final solutions was brought to the values of 0·04 per cent., 0·02 per cent., 0·01 per cent., 0·005 per cent., 0·0025 per cent., 0 per cent. After one night in the cold room, each sample was divided in half. To the first of these 5-ml. samples, 0·6 ml. of 1 M KC1 was added, and to the second 0 6 ml. of 30 per cent. urea. Following the methods of Ranzi and Citterio in the paper cited, the viscosities were read 4 hours later. The values obtained are shown in Table 1. The effect of trypan blue in increasing the viscosity, and in rendering the protein particles more resistant to the action of urea, is very clearly seen. In fact, if the data are plotted (Text-fig. 1), they exhibit a picture very similar to that found for lithium and shown in fig. 9 of Ranzi & Citterio’s paper. There is, then, clear evidence that trypan blue falls into line with what would be expected from it on Ranzi’s hypothesis. This can probably be taken as rather strong support for the hypothesis, since trypan blue is a substance of quite different chemical nature from the other vegetalizing agents known.

Table 1

Specific viscosities of euglobulin. Determined by Prof. S. Ranzi, Milan, by the methods described in Ranzi & Catterio (1954)

Specific viscosities of euglobulin. Determined by Prof. S. Ranzi, Milan, by the methods described in Ranzi & Catterio (1954)
Specific viscosities of euglobulin. Determined by Prof. S. Ranzi, Milan, by the methods described in Ranzi & Catterio (1954)
Text-fig. 1.

Specific viscosities of solutions of euglobin a and b from frogs’ eggs, diluted either with KC1 or urea, and containing different percentages of trypan blue. Determinations by Prof. S. Ranzi.

Text-fig. 1.

Specific viscosities of solutions of euglobin a and b from frogs’ eggs, diluted either with KC1 or urea, and containing different percentages of trypan blue. Determinations by Prof. S. Ranzi.

Although this investigation by Professor Ranzi gives us some basis for understanding the action of trypan blue in causing mesodermalization in the amphibian egg, it must be remembered that the substance has several other effects which have been enumerated above. It is not possible at present to see how these can all be brought under the same heading as the mesodermalization. It is possible, indeed, that the morphogenetic effects on gastrulation and the elongation of the tail-bud may also be the results of an influence on the physical state of the protein particles in the cytoplasm, but this at present remains mere speculation. The disturbance of the fluid balance leading to oedema, the effects on the epidermis, and the vacuolar degeneration of neural and mesodermal tissues, are probably to be attributed to more general toxic properties of trypan blue, not necessarily immediately related to its influence on protein viscosity.

It will be noted that there is some difference in the incidence of the various types of abnormalities in the different species of amphibia used. Thus mesodermalization was common in Xenopus and in the small series of Axolotl but was not seen in Triturus species, which showed only rather unspecific effects suggestive of a general toxicity. A comparison of the results of treating amphibian and mammalian embryos also provides evidence that the genetic nature of the material strongly affects the kind of abnormalities produced. In the latter group, few signs have been noticed of a mesodermalization of the notochord, although Murakami, Kameyama, & Kato (1954) have described some embryos which might perhaps be interpreted in this way. Hamburgh (1954) mentions what he calls ‘wavy notochords’, but this abnormality seems to be due to a general disturbance of growth of the somites rather than to a mesodermalization of the chorda. The production of oedema accompanied by blistering is, however, an effect which is as marked among the mammals as in the amphibia. Waddington & Carter (1953), indeed, suggested that most of the abnormalities in mouse embryos of their strain could be interpreted as secondary results of such disturbances of the fluid balance. In particular, the abnormalities in the tail region seemed often to result from the formation of local accumulations of blood or body fluids. In so far as this is the case, they would not be truly comparable to the abnormalities of the tail-bud region described in the Amphibia, since these appear to be due to a direct effect on the morphogenetic movements. It may be, however, that similar direct effects played a greater part in causing the tail abnormalities of mouse embryos than was previously thought.

There are considerable differences in the effects of trypan blue on the nervous systems of the mammals and the amphibia. In the latter, we have seen striking examples of the production of microcephaly, and in general the effect of the dye is to bring about a reduction of the nervous system. In mammals all workers agree that there are considerable abnormalities in the closure of the neural folds in the period shortly after the administration of the dye. In Waddington & Carter’s material (CBA inbred mice) it appeared that embryos affected in this way usually died at the age of about 12 to 13 days. In other stocks many of them survive to later stages. Both Hamburgh and the Japanese workers point out that in such surviving embryos there may be a considerable hypertrophy of the nervous system, involving overgrowth rather than reduction. In some cases, the final result at birth was a typical pseudencephaly, with an enlarged and unenclosed brain sitting on top of the head like a cap. The frequency of this abnormality seems to vary in different mouse stocks. It amounted to something under 18 per cent, in the Bagg albino strain used by Hamburgh, to about 7 per cent, in an albino strain used by Murakami, to less than 1 per cent, in certain pigmented stocks used by the same author, while it was not seen at all in the CBA’s used by Waddington & Carter. It also appeared in the rats studied by Gillman et al. but is not described in the fairly small series of abnormalities produced by Harm (1954) in the rabbit.

Gillman
,
J.
,
Gilbert
,
C.
,
Gillman
,
T.
, &
Spence
,
I.
(
1948
).
A preliminary report on hydro-‘cephalus, spina bifida, and other congenital anomalies in the rat produced by trypan blue
.
S. Afr. J. med. Sci
.
13
,
47
90
.
Gustafson
,
T.
(
1950
).
Survey of the morphogenetic action of the lithium ion and the chemical basis of its action
.
Rev. suisse Zool
.
57
,
77
92
.
Hadorn
,
E.
(
1951
).
Experimenten bewirkte Blockierung der histologischen Differenzierung in der Chorda von Triton
.
Arch. EntwMech. Org
.
144
,
491
520
.
Hamburgh
,
M.
(
1952
).
Malformations in mouse embryos induced by trypan blue
.
Nature, Lond
.
169
,
27
.
Hamburgh
,
M.
(
1954
).
The embryology of trypan blue induced abnormalities in mice
.
Anat. Rec
.
119
,
409
22
.