The uptake of water by haploid and diploid sibling embryos of Xenopus laevis has been investigated by measuring the density changes which occur during the development of intact embryos from the blastula to the late tail-bud stage, and of explants from which most of the presumptive endoderm has been removed.

The results show that up to the mid-gastrula stage there is no difference between the haploid and diploid embryos; but from then on, whereas the diploid volume increases steadily, the haploid gastrulae undergo a series of cyclical volume changes due to loss of fluid through the blastopore. It is concluded that this is the result of an excessive inflow of water through the haploid ectoderm, because it was found that the volume of haploid ectodermal explants increased much more rapidly than the volume of similar diploid explants. Excess flow through the haploid ectoderm also accounts for other characteristics of the haploid syndrome-microcephaly and lordosis.

It is suggested that it is the doubling of the cell number in haploid embryos with the consequent 25% increase in aggregate cell membrane area which accounts for the difference between the uptake of water by the two types of embryos. It is also suggested that changes in the rate of water flow through the ectoderm and endoderm which are thought to account for the accumulation of water in the blastocoel and archenteron in the normal diploid embryo arise in a similar way.

The majority of haploid embryos of Xenopus laevis develop into larvae which show a characteristic syndrome -large accumulations of water occur under the skin, the nerve cord is short, the brain is small, and there is considerable lordosis. A study of the very few haploid embryos which do not show this syndrome led Fox and Hamilton to conclude that the oedema is a result of an excessive inflow of water through the ectoderm rather than renal failure, because they found that in these embryos the kidneys were hypertrophied (Fox & Hamilton, 1964).

Accumulation of water in intercellular spaces, however, is a normal feature of early development in the diploid embryo; water accumulates at an increasing rate first in the blastocoel, a development of the cleavage cavity, and then in the archenteron -a cavity formed by an invagination of the blastula surface and lined by cells derived from the vegetal pole of the blastula. During the formation of the archenteron the blastocoel decreases in volume and disappears.

It is therefore of some interest to know whether the factors which give rise to the abnormal inflow of water through the haploid larval skin also affect the uptake and distribution of water in the earlier stages of development. We have accordingly carried out a series of experiments to determine the rate of water uptake by decapsulated diploid and haploid siblings from the early blastula to the late neurula stage, and also the water uptake of vesicles formed by blastula explants from which most of the presumptive endoderm has been removed and in which gastrulation does not occur. The results of these experiments indicate that the rate of transcellular water flow in the presumptive ectoderm and endoderm of the haploid embryo is abnormally high from the gastrula stage onwards but only gives rise to abnormal volume changes in intact embryos during the neurula stage.

Embryos

The eggs were obtained from adult Xenopus laevis which were induced to spawn by injection with chorionic gonadotrophin. The eggs were collected at 5 min intervals, and 10 min after laying half the eggs were irradiated with u.v. light (2 × 10−4 J mm−2) to produce androgenetic haploids (Gurdon, 1960). The treated and control embryos were kept in crystallizing dishes in sterile tap-water at 25°C in a thermostatic water bath. In what follows we shall assume that irradiated embryos are haploid and we shall refer to them as such.

Density gradients

Linear density gradients of colloidal thorium oxide stabilized with dextrin were made by upward displacement in parallel-sided glass tubes using a simple gradient machine; the gradient tubes were then transferred to a glass-fronted thermostatic bath at 25°C (±0·001°C). Details of the gradients are given in Table 1.

Table 1

Details of typical density gradient

Details of typical density gradient
Details of typical density gradient

The gradients were calibrated by measuring the position of glass density standards that had previously been calibrated against droplets of KC1 solutions in a brom-benzene/kerosene (Solve Esso 15) gradient set-up in a similar way and saturated with water as described by Linderstrom-Lang & Lanz (1938). Using this technique it is possible to make gradients 20 cm high with a density range from top to bottom of 0·045 gcm−3givingasensitivityof0-0002gcm−3mm−1 and linearity such that changes in density can be measured to ± 0·0001 g cm−3. In experiments in which the embryos remained in the gradient throughout their development, the oxygen tension was maintained by circulating oxygen or air through loops of thin-walled polyethylene catheter immersed in the gradients.

Density measurements

In order to obtain the density data required to calculate the change in volume of the haploid and diploid embryos three series of experiments were carried out. In the first, successive batches of control and irradiated sibling embryos of known age were decapsulated surgically in l/10th Holtfreter solution and their density measured 10 min after they had been introduced into the density gradient.

In the second series, four or five haploid and diploid blastulae were placed in identical gradients set up side by side in the thermostat bath. The gradients were photographed at intervals of 15 min with an automatic ‘Robot’ camera. The density of the embryos was determined by projecting the negative image of the gradients on to a linear density scale using the calibration beads as reference points.

In a third series of experiments the same technique was used to monitor the density change of ectodermal explants from irradiated and unirradiated embryos. In these experiments, the bottom third of stage 8 (Nieuwkoop & Faber, 1963) sibling blastulae was removed with fine forceps. The remaining upper two thirds were allowed to round up ini/10th Holtfreter solution before being introduced into the gradients. These explants formed stable vesicles whereas similar explants from a stage 9 blastula containing more ectoderm tended to burst.

The determination of changes in volume and water content

It can be shown that the volume changes which occur during Xenopus development up to the early tail-bud stage are entirely due to changes in the water content of the embryo as follows: the weight of the embryo in water, its reduced weight (RWe)is a function of the mass of the embryo (We), its volume (Ve)and the density of water at the same temperature (pw)
formula
The density of the embryo (pe)is in turn a function of its mass and volume:
formula
From (1) and (2) the reduced weight can be expressed in terms of its mass and density,
formula
or as the sum of the reduced weights of its constituents (Σ RWi) and their weights (Wi)and densities (pi) (Løvtrop, 1953):
formula

Since the reduced weight remains unchanged from cleavage to the early tailbud stage (Tuft, 1962) it follows from (4) that any change in density must be due to the uptake or loss of matter with the same density as water.

The dry mass of the embryo would tend to decrease during development as the result of a loss of solutes and the oxidation of respiratory substrates. The former is so small that it cannot be detected (unpublished data) and the latter, which is also small, can be estimated from the oxygen consumption of the embryo. The rate of oxygen consumption rises gradually during development of the Xenopus embryo, reaching 4·8 × 10−4ml h−1 per embryo during the late neurula stage (Tuft, 1953). If we assume that 1 g of respiratory substrate (p = T000) involves the uptake of 1·0 × 103 ml O2 then weight will be lost at the rate of 4·8 × 10−4 mg h−1; thus for an embryo with a density of 1·050 g cm−3 and reduced weight of 0·0879 mg, the density will increase at a rate of 1·4 × 10−5g cm−3 h−1. This is below the limit of resolution of the density measuring technique we have used.

The relative change in volume of the embryo or its water content (Vt/V0) can then be calculated from the density of the embryo at t = 0 (p0) and t (pt) and the density of water (pw) as follows.

From equations (1) and (2):
formula
formula
formula

Morphological differences

Examples of the morphological differences between haploid and diploid embryos from the blastula stage to late gastrula are illustrated in Fig. 1. As will be seen, there is very little difference between the anatomical appearance of the two kinds of embryo except that onset of gastrulation is delayed in the haploids and, by the time it does begin, the haploid blastocoel is very much larger than that in the corresponding diploid embryo. The preparations also show that the dorsal lip of the blastopore is less tightly applied to the yolk plug and that the archenteron contains less fluid in haploids.

Fig. 1

Drawings of half embryos arranged to illustrate the difference between the rate of morphological development in haploid and diploid embryos of Xenopus laevis. Each haploid embryo in the left-hand column was the same age as the corresponding diploid embryo in the right. Aa =Archenteron, Bc = blastocoel.

Fig. 1

Drawings of half embryos arranged to illustrate the difference between the rate of morphological development in haploid and diploid embryos of Xenopus laevis. Each haploid embryo in the left-hand column was the same age as the corresponding diploid embryo in the right. Aa =Archenteron, Bc = blastocoel.

Density changes

Intact embryos

The results of the first series of experiments, in which the densities of embryos in a series of different age-groups were measured, showed that up to the late blastula stage there was no significant difference between the irradiated and control groups, but during the late gastrula and neurula stages the two groups differed considerably. The densities of the irradiated group were more variable than the control group and had a significantly higher mean density. However, after the control embryos had collapsed and the archenteron had emptied, the difference between the mean densities was again insignificant.

In the second series of experiments the embryos were allowed to remain in the gradient throughout their development and the density was monitored at 15 min intervals. The results of one such experiment are is shown in Fig. 2, where the density changes which took place in each embryo are shown. It will be seen that, as before, the density of both haploid and diploid embryos decreased uniformly until the late gastrula stage. The diploid densities then continued to decrease steadily until the archenteron collapsed at 22 h, except for a brief transient increase at stage 13 when the yolk plug is withdrawn. The density of the haploid embryos on the other hand went through a series of cyclical changes of varying amplitude which lasted from the mid-gastrula to the late neurula stage.

Fig. 2

The density of four haploid and four diploid embryos developing in a density gradient at 25°C, measured at 15 min intervals, plotted against age.

Fig. 2

The density of four haploid and four diploid embryos developing in a density gradient at 25°C, measured at 15 min intervals, plotted against age.

Results from all experiments in the second series have been pooled and the mean density of all embryos of the same age has been calculated. The samples were tested for homogeneity, and where the F value at the 5% level was not significant the difference between the mean values of treated (androgenetic haploids) and control groups (diploids) was tested using a two tailed t test. When the variances of the samples were not homogeneous, a non-parametric test was used.

The results of this analysis given in Table 2 confirm the results of the earlier experiments; they also show that from 24 h to 48 h the mean density of haploid and diploid embryos does not differ significantly. The relative volume changes (Vt/V0) calculated from the mean density values in Table 2 are shown in Fig. 3.

Table 2

The difference between the density of diploid and haploid embryos at different ages

The difference between the density of diploid and haploid embryos at different ages
The difference between the density of diploid and haploid embryos at different ages
Fig. 3

The relative volume change (Vt/V10) of haploid and diploid embryos from the age of 10 to 27 h at 25°C calculated from the mean densities in Table 2.

Fig. 3

The relative volume change (Vt/V10) of haploid and diploid embryos from the age of 10 to 27 h at 25°C calculated from the mean densities in Table 2.

Open embryos

Density measurements on embryos which have been operated on in such a way that the blastocoel and archenteron are open to the environment (‘opened embryos’) show that at the early and mid-gastrula stage the mean density of the haploid cell mass is significantly less than that of the corresponding diploid cell mass. The difference, however, is small and represents an increase in volume of about 4% (Table 3). Measurements made at later stages suggest that this difference does not persist.

Table 3
graphic
graphic

Animal pole explants

Attempts to compare the volume changes of vesicles made from the roof of the late blastulae failed because, although they formed vesicles, they were un-stable, going through a series of cyclical density changes and finally disintegrating.

However, when stage 8 embryos were used and only the vegetal third of the blastula was removed -that is to say, most of the presumptive endoderm -the vesicles were more stable and behaved very much like normal embryos except that they did not gastrulate.

The relative volume changes Vt/V0 of explants of the latter type calculated from equation (7) are plotted against time in Fig. 4. This shows that the relative volume of haploid explants increases more rapidly and they reach their maximum sooner than the diploids.

Fig. 4

Changes in the relative volume (Vt/V0) of vesicles formed by blastula explants from which most of the vegetal pole material has been removed. Calculated from the density of each vesicle measured at 15 min intervals after it was placed in the density gradient.

Fig. 4

Changes in the relative volume (Vt/V0) of vesicles formed by blastula explants from which most of the vegetal pole material has been removed. Calculated from the density of each vesicle measured at 15 min intervals after it was placed in the density gradient.

The first thing to note about the results obtained in the present experiments is that the density/time curves for the diploid embryos differ from those previously published (Tuft, 1962, 1964). They do not show a decrease in rate of density change during gastrula stages and the collapse of the late neurula occurs 22 h rather than 18 h after fertilization. These differences are the result of the way in which the data for the density/time curves were obtained in the two series of experiments.

In the earlier experiments embryos were decapsulated surgically immediately before their density was measured and the curves constructed from the mean densities of the different age-groups. In the present experiments, on the other hand, the curves are based on successive measurements of the density of individual embryos after decapsulation at the late blastula stage.

Deformation of the embryo during removal of the capsule at the gastrula stage is responsible for the apparent decrease in the rate of density change in the earlier experiments because it tends to cause a loss of fluid from the newly formed archenteron before the blastopore is tightly closed. This tends to increase the mean density of embryos at this stage. At later stages when the blastopore is tightly closed loss of fluid is less likely to occur.

In encapsulated embryos the elasticity of the capsule opposes the elongation of the notochord and long axis of the embryo at the late neurula stage. As a result, when the embryo loses its lateral stability it jackknifes, causing a sudden and complete emptying of the archenteron cavity. In decapsulated embryos, on the other hand, there are no external forces acting on the embryo, and under these circumstances the only forces tending to raise the pressure inside the archenteron are elastic forces developed in the body wall itself as the embryo elongates. These forces take longer to develop sufficient pressure to collapse the archenteron.

Nevertheless, monitoring the density of naked embryos after decapsulation at the blastula stage reveal important differences in the uptake of water by haploid and diploid embryos. The density/time curves of the kind illustrated in Fig. 2, for example, show that from the blastula until the mid-gastrula stage, density changes in the two types of embryo do not differ significantly, and this is confirmed by the combined results in Table 2. The net accumulation of water must therefore be the same in both types of embryo (see above). This does not necessarily mean, however, that its distribution within the embryo is the same. From Table 3 it will be seen that the density of the cell mass in haploid embryos tends to be slightly less than in the diploid -that is, its volume is greater and the volume of the large intercellular spaces smaller. The morphological data, on the other hand, suggest that when gastrulation begins in the haploid (Fig. 1) the blastocoel is in fact enlarged. These two observations are not incompatible, because when haploids reach this stage the archenteron in the diploids has already begun to form, and it is not possible to distinguish between density changes due to water accumulating in the blastocoel and in the archenteron from the density of the intact embryo alone.

From the mid-gastrula stage onwards, however, the two types of embryo behave very differently (Table 2). Whereas the mean density of the haploid embryos remains more or less constant, the diploid density decreases continuously from 14 to 23 h. The density/time curves of individual embryos (Fig. 2) show that the difference between the two types of embryo is even more striking; the haploid neurulae, unlike the diploid neurulae, undergo a series of cyclical density changes which are the result of successive filling and emptying of their archenterons. After 23 h, however, there is again no significant difference between mean density of the haploid and diploid embryos; that is to say, there is no difference between their volumes. This also appears to be true for subsequent stages in spite of the fact that haploid embryos look very abnormal, with large water-filled cavities under the skin. However, the procedures used at these stages to immobilize the embryos may alter their water content. For this reason we will only consider the differences between haploid and diploid embryos in the pre-collapse stages.

For reasons that have been given in earlier papers (Tuft, 1962, 1964) it has been suggested that water-regulating mechanisms in the cell membranes maintain the relatively constant cell volume observed during the early stages of development, and are so arranged that they also give rise to a net transcellular inflow of water through the animal pole and a net outflow through the vegetal pole of the blastula. The difference between the magnitudes of these two flows results in an accumulation of water in the blastocoel. But when, subsequently, the derivatives of the vegetal pole cells come to line the archenteron cavity, as a result of invagination, both flows are directed inwards and give rise to a very rapid increase in the volume of the archenteron cavity (Tuft, 1962). This is illustrated diagrammatically in Fig. 5.

Fig. 5

Diagram to illustrate the net water flows through presumptive ectoderm and endoderm in the Xenopus embryo at different stages in its development after Tuft (Tuft, 1962). Aa = Archenteron; Bc = blastocoel.

Fig. 5

Diagram to illustrate the net water flows through presumptive ectoderm and endoderm in the Xenopus embryo at different stages in its development after Tuft (Tuft, 1962). Aa = Archenteron; Bc = blastocoel.

If this hypothesis is correct then the difference between the behaviour of the haploid and diploid embryos could be explained in one of two ways: either the blastopore lips in the haploid neurulae are weakened in some way and are unable to withstand the normal pressures developed within the archenteron by a normal inflow of water, or alternatively the net inflow across the haploid ectoderm and endoderm is abnormally high and the blastopore lips cannot withstand the excess pressure developed.

The behaviour of animal pole explants enables us to distinguish between these two alternative explanations. It will be recalled that in these experiments most of the vegetal (that is, endodermal) surface of the blastula was removed and the remaining portion, comprising mainly presumptive ectoderm, was allowed to round up. It will be seen from the results illustrated in Fig. 4 that the volume of vesicles formed by haploid explants increased very much more rapidly than the volume of similar diploid vesicles, and after reaching a maximum also decrease more rapidly, suggesting that both the inflow and outflow are increased.

The decrease in volume is not due to bursting of the vesicle, because the rate of change is too small, but probably results from an increase in the net outflow of water due to the growth of vegetal pole material left in the explants when they were made. This is consistent with the observation that diploid vesicles take about 6 h to empty whereas in intact diploid embryos, which have more vegetal pole surface, the blastocoel empties in 3 h.

These experiments demonstrate that the flow through the haploid ectoderm is greater than that through the diploid, and it follows from what has been said earlier that the flow through the haploid endoderm is also greater. We can con-elude therefore that the successive filling and emptying of the intact haploid archenteron is at least in part the result of an increased inflow of water.

It is interesting to note that an increased inflow through the ectoderm could also account for dorsal flexure and microcephaly -two other characteristics of the haploid syndrome. The neural tube is formed by invagination of the neur-ectoderm, and in the normal diploid embryo it is flexed ventrally at first, but as water is removed from its lumen it straightens out. An increase in this outflow would therefore be expected to maintain a very much reduced neural volume.

The haploid neurulae differ from the diploid in one other important respect -they have twice as many cells, and the cells are half the size of those in the diploid. Although the aggregate cell volume is the same in both haploid and diploid embryos, doubling the cell number gives rise to an increase of 25% in the aggregate cell surface area in the haploids. (See Appendix.)

If the hypothesis outlined earlier is correct, an increase in cell number will therefore increase the number of water-regulating sites responsible for the net transcellular water flows across the ectodermal and endodermal cells.

The explant experiments lend support to this view. It will be recalled that the explants used in these experiments, which consisted of all the presumptive ectoderm and a little of the endoderm, were made at the mid-blastula stage and were allowed to heal before being introduced into the density gradient. That is to say the measurements began when the intact embryo would normally be at stage 9. At this stage the rate of cell multiplication in diploid embryos decreases rapidly until it reaches the relatively low rate characteristic of later stages in development. In haploid embryos on the other hand the high rate of division characteristic of the earlier stages is maintained for about an hour until the cell number has been doubled, after which the rate becomes the same as the diploid.

The aggregate cell surface in the haploid explants will therefore tend to increase very much more rapidly than it does in the diploid, and, if our hypothesis is correct, the explants should reach their maximum rate of volume change sooner than the diploids, which, as we have seen, is exactly what the experimental results show (Fig. 4).

We may therefore conclude from these experiments that the abnormal development of the haploid embryo, like that of the tadpole described by Fox & Hamilton (1964), is the result of an excessive inflow of water. Furthermore if, as has been suggested, the uptake of water by the embryo involves water-regulating sites in the cell membranes, the excess flow into the haploid archenteron can be attributed to the increase in total cell membrane area which is a consequence of doubling the number and halving the size of the cells at the late blastula stage.

Changes in the relative rate of water flow through the presumptive ectoderm and endoderm which are thought to account for the accumulation of water in the blastocoel and archenteron of the normal diploid embryo may arise in a similar way. Thus during the formation of the blastocoel the rate of cell division in the presumptive ectoderm is very much greater than it is in the endoderm, whereas the reverse is true during gastrulation and formation of the archenteron. If this is so the simple model put forward to account for the uptake and distribution of water by the Xenopus embryo in earlier papers (Tuft, 1962, 1964) will have to be modified and detailed information obtained about the ultrastructure and dimensions of the cells in the two layers.

APPENDIX 1

Effect of cell number on total area of cell membrane

The aggregate cell surface area A in an embryo consisting of n1 cells and having a total cell volume V can be calculated as follows: let v1 r1a1 be the volume, radius, and area respectively of the individual cells. Then assuming the cells are spherical,
formula
formula
formula
Then the aggregate surface area
formula
and from (3) and (4)
formula
Similarly, for an embryo with the same volume V but with n2 cells,
formula
The ratio of the aggregate cell surface area in the two embryos from (5) and (6) is then
formula
For diploid and haploid embryos where Vhap = Vhap and Vhap = 2ndlp,
formula

This work was carried out in the Department of Zoology, University of Edinburgh. We would like to thank Professor Mitchison for the facilities provided, Mr Holmes who did the illustrations, and Mrs Ann Muir for technical assistance. One of us, P. H. Tuft, also wishes to thank the Distillers Company for the grant in aid.

Fox
,
H.
&
Hamilton
,
L.
(
1964
).
Pronephric system in haploid and diploid larvae of Xenopus laevis
.
Experientia
20
,
289
.
Gurdon
,
J. B.
(
1960
).
The effects of ultraviolet irradiation on uncleaved eggs of Xenopus laevis
.
Q. JI Microsc. Sci
.
101
,
299
311
.
Linderstrom-Lang
,
K.
&
Lanz
,
H.
(
1938
).
Studies on enzymatic histochemistry. XXIX. Dilatometric micro-estimation of peptidase activity
.
C. r. Trav. Lab. Carlsberg, (Chim.)
21
,
315
338
.
Lovtrop
,
S.
(
1953
).
Energy sources of amphibian embryogenesis
.
C. r. Trav. Lab. Carlsberg
28
, no.
14
,
372
396
.
Nieuwkoop
,
P. D.
&
Faber
,
J.
(ed.) (
1956
).
Normal Table of Xenopus laevis (Daudini)
.
Amsterdam
:
North Holland Publishing Co
.
Tuft
,
P. H.
(
1953
).
Energy changes in development
.
Archs. néerl. Zool
.
10
, (
Suppl 1
),
59
75
.
Tuft
,
P. H.
(
1962
).
The uptake and distribution of water in the embryo of Xenopus laevis
.
J. exp. Biol
.
39
,
1
19
.
Tuft
,
P. H.
(
1964
).
The uptake and distribution of water in the developing amphibian embryo
.
Symp. Soc. exp. Biol
. no.
xix
, pp.
385
402
.
Cambridge University Press
.