/. Embryol. exp. Morph. Vol. 33, 1, pp. 57-74, 1975
Printed in Great Britain
57
Quantitative studies of germ plasm and germ cells
during early embryogenesis of Xenopus laevis
By P. McD. WHITINGTON1 AND K. E. DIXON2
From The School of Biological Sciences, The Flinders University
of South Australia
SUMMARY
The germ plasm in the egg is partitioned between the first four blastomeres by thefirsttwo
cleavage planes. Although the blastomeres divide 10-11 times through the rest of cleavage,
as shown by reduction in their size, the number of presumptive primordial germ cells (p.p.
germ cells) does not increase significantly. During and as a result of the formation of the
first two cleavage planes, the germ plasm aggregates together and moves towards and along
the cleavage furrows. At subsequent mitoses, the germ plasm is localized at one of the poles
of the spindle and hence is segregated to only one of the daughter cells, thus explaining how
mitosis occurs without increase in the number of cells with germ plasm. Early in gastrulation,
the germ plasm moves to a perinuclear position, therefore ensuring that as mitosis continues,
both daughter cells receive germ plasm and the number of p.p. germ cells increases. Direct
counts of the number of p.p. germ cells and measurements of their volume suggest that they
divide twice between early gastrula and the stage at which they leave the endoderm. The p.p.
germ cells behave similarly to the adjacent endodermal cells until they begin to migrate to
the gonad, an event which may represent the first overt signs of differentiation. Measurements
of the volume of germ plasm suggest that there is no change through cleavage. The general
conclusion is drawn that during cleavage, the morphogenetic determinant germ plasm is
segregated to a few cells by the normal processes of cleavage and that subsequently these
cells undergo a small number of cloning divisions which are contemporaneous with the first
signs of differentiation.
INTRODUCTION
In some anuran amphibians, primordial germ cells, i.e. those from which the
gametes will be derived, can be distinguished from somatic cells throughout
embryogenesis by the presence of a microscopically identifiable cytoplasmic
inclusion known as 'germplasm' (see reviews by Bounoure, 1934,1964; Bladder,
1965, 1966, 1970). Descriptive studies have shown that the germ plasm is localized in the vegetal pole cytoplasm and hence the germ cells are located initially
in the presumptive endoderm. Later in embryogenesis the germ cells are reported
to migrate from the endoderm into the genital ridges, there to proliferate and
then enter meiosis. Most authors use the term 'primordial germ cells' without
1
Author's address: Research School of Biological Sciences, The Australian National
University, Canberra, A.C.T. 2600, Australia.
2
Author's address: School of Biological Sciences, The Flinders University of South
Australia, Bedford Park, S.A. 5042, Australia.
58
P. MCD. WHITINGTON AND K. E. DIXON
intending to specify whether the cells have reached the gonad or not. We suggest
that slightly more precision in terminology is required and therefore we will use
the term' presumptive primordial germ cells' (p.p. germ cells) when it is necessary
to indicate specifically primordial germ cells which have not reached the gonad,
and restrict the term 'primordial germ cells' to those germ cells which have
reached the gonad and which will give rise eventually to either spermatogonia
or oogonia by mitosis (see also Kerr & Dixon, 1974).
The descriptive studies have been complemented by experimental work which
provides strong evidence that germ plasm plays an important role in defining
germ line cells (see, for example, Smith, 1966; Buehr & Blackler, 1970).
However, the actual mechanism by which germ plasm specifies germ line cells
remains unresolved. Not the least of the problems associated with defining the
mechanism of action of germ plasm is that, since it is, of necessity, always
present (Weismann, 1885), it has a potential for action throughout the whole
life-cycle. Our work has been directed, in part, towards specifying the time when
germ plasm is active, as an initial step towards defining its mechanism of action.
This paper reports the results of a quantitative study of the number, size and
position of p.p. germ cells and of the amount and intracellular position of germ
plasm in embryos of Xenopus laevis from the first cleavage divisions up to the
time the p.p. germ cells leave the endoderm to enter the genital ridge.
MATERIALS AND METHODS
Ovulation was induced in mature female X. laevis by injecting chorionic
gonadotrophic hormone (Pregnyl, British Pharmaceuticals). Eggs were manually
stripped from the female and fertilized by exposing them to a macerated testis
(after Wolf & Hedrick, 1971). Embryos were staged according to the normal
tables of Nieuwkoop & Faber (1967).
To facilitate fixation and infiltration, jelly membranes surrounding the egg
were removed by treating for 3-5 min with a 1 % cystein-2 % papain solution
at pH 8-0, then washing thoroughly in distilled water.
Eggs and embryos were fixed in Smith's fixative (after McClung, 1964) overnight, then washed in distilled water for 12-24 h. Embryos were dehydrated in
alcohol and cleared in xylene, or dehydrated and cleared in dioxan (Puckett
in McClung, 1964). All material was embedded in paraffin (Paraplast) and
sectioned at 5 /tm.
Sections were stained with Heidenhain's Azan (after Gurr, 1962) or Janus
Green-Neutral Red (after McClung, 1964) or Acid Fuchsin-Azur II (Volkonsky, 1928). The Azan stain permitted a more positive identification of germ
plasm but the other stains were particularly useful for demonstrating cell and
nuclear membranes. Measurements were taken with the aid of a Zeiss integrating
and grain-size disc. Germ plasm volume was estimated by measuring the area
of the patch of germ plasm in successive sections then multiplying the sum of
Germ plasm and germ cells in Xenopus
59
Table 1. Number of cells containing germ plasm
Stage
4-cell to morula
Blastula
Gastrula
Early tail-bud
Late tail-bud
End of endodermal phase
Stage no.
3-7
8,9
11-13
19,20
29,30
38-41
No. of
embryos
No. of cells
(M± S.D.)
13
11
9
7
6
4-5 ±0-9
5-3 ±1-9
8-7 ±3-0
7-4±4-l
9-2±3-5
13-9±4-8
11
these areas by the total thickness of the sections in which the germ plasm was
present. Germ-cell volume was estimated in a similar manner except that
measurements were taken only on every fifth or tenth section, depending on the
size of the cells.
RESULTS
Number of germ cells
The number of cells which contain inclusions of germ plasm (and hence
identified as presumptive primordial germ cells) in embryos at different stages
of development is shown in Table 1. At the four-cell stage, each of the blastomeres contains germ plasm, and throughout cleavage there is no significant
change in this number.
As the germ plasm initially lies more or less symmetrically around the vegetal
pole (see Fig. 6A, B), the first two cleavage planes must divide the germ plasm
between the first four blastomeres. Direct observation of both of the fourth
cleavage planes showed that they passed exactly through the point of intersection of the first and second cleavage planes in only 5 % of eggs approximately;
in the remainder of the eggs examined, the cleavage pattern was as indicated in
Fig. 1. This pattern of cleavage provides a mechanism whereby the germ plasm
remains in just four blastomeres at the 32-cell stage (see discussion of the intracellular sites of germ plasm below). The observation that the number of cells
which contain germ plasm does not increase in later cleavage (i.e. after the 32-cell
stage) does not imply that they do not divide. On the contrary, reduction in the
size of the cells indicates that division continues, but since the germ plasm is
located at one of the spindle poles it is distributed to only one of the daughter
blastomeres at each cleavage (see section on intracellular sites of germ plasm).
Between formation of the blastula and stages 38—41 when the p.p. germ cells
leave the endoderm, the average number increases from 5 to 14, suggesting
that one to two divisions of each germ cell (on the average) take place during
this time.
In intermediate developmental stages (gastrula, early and late tail-bud stages)
the number of germ cells observed (Fig. 2) suggests that in a small proportion
60
P. MCD. W H I T I N G T O N AND K. E. DIXON
Fig. 1. Diagrammatic representation of the position of the 4th cleavage planes
(dotted lines) relative to the vegetal pole. The pattern represented in (A) occurred
in approximately 70 % of the eggs examined, in (B) in about 20 % of the eggs, and in
(C) in about 5 %. In the remaining 5 % of eggs, the 1st, 2nd and 4th cleavage planes
intersected at the one point (D).
of the embryos, both mitoses take place early, at about the gastrula stage, but
more usually at least one of the mitoses occurs between stages 29, 30 and 38-41.
Size of germ cells
During the developmental period studied the volume of the individual germ
cells decreases (Table 2). Reduction in cell size may be due partly to utilization
of yolk but cell division will be a much more important factor. Although there
are large variations in the volume of cells at any one developmental stage, the
data taken overall indicate that division occurs throughout cleavage and later
embryogenesis. The reduction in size suggests that between the 8-cell stage
and gastrulation about 10-11 mitoses take place, although, as shown earlier,
there is no increase in number of p.p. germ cells. Between gastrula and stage
38-41 when the p.p. germ cells leave the endoderm, two cloning divisions occur,
a finding which is consistent with the conclusions reached from the direct counts
of cell numbers.
Germ plasm and germ cells in Xenopus
61
8-cell to morula
Blastulu
Gastrula
Stage 19, 20
Stage 29, 30
Stage 38-41
4
8
12
16
Number of germ cells/embryo
Fig. 2. Frequency distribution of the numbers of presumptive primordial germ cells
in embryos at different stages of development from early cleavage to early larval
forms in which the germ cells migrate from the endoderm. There is no significant
change in number of cells during cleavage, i.e. between 8-cell to morula and blastula.
Between blastula and stage 38-41, the increase in the number of cells in each embryo
suggests that about two divisions of each cell have taken place. The numbers of cells
at gastrula, stages 19, 20 and 29, 30, suggest that in a small number of embryos the
divisions occur about gastrula, but in most embryos the increase in the number of
germ cells occurs after stage 29, 30.
Position of germ cells
In the earliest cleavage stages, the p.p. germ cells lie at the vegetal pole (Fig.
3 A). In the blastula, the p.p. germ cells are found above and around the vegetal
pole in the bottom third of the embryo (Fig. 3B). The slight displacements from
their original position can be attributed to shuffling of cells during cleavage.
In the early gastrula the p.p. germ cells occur within the involuting endoderm
(Fig. 3 C) and in the completed gastrula they lie in the floor of the archenteron
towards the blastopore (Fig. 3D). Their final resting place and their position at
intermediate stages of gastrulation suggest that the p.p. germ cells are carried
passively with the surrounding endoderm and participate as endodermal cells
do in the morphogenetic movements of gastrulation.
62
P. MCD. WHITINGTON AND K. E. DIXON
Table 2. Volume of germ cells (jam3 x V04)
Stage
8 to 16-cell
32-cell
Morula
Early-mid-blastula
Blastula
Early-mid-gastrula
Early tadpole
Stage no.
4,5
6
6i
7,8
9
101-12
38-41
No. of
embryos
No. of cells
6
5
2
5
4
4
7
10
10
6
11
11
13
47
Volume
(M±S.D.)
5583 ±1086
3390 ±928
1263 ±327
343 ±139
103 ±46
4-6±l-8
l-3±01
At stages 20 and 29, 30, the p.p. germ cells occupy the same relative position
within the embryo as shown by measurements of their distance from the ventral
margin of the endoderm (Table 3; see also Fig. 3E, F). By stage 38 the position
of the germ cells has changed dramatically and they are found at the top of the
endoderm. The change in position appears to be due to active movements of
the p.p. germ cells since there are no general shifts known in the tissues of the
embryo at this time by which the germ cells could be passively transported to
their new location. Tahara & Nakamura (1961) and Harris (1967) have suggested that very late in development of some urodeles and anurans there is a
reorganization of the endoderm to form the small intestine (but see also Ballard,
1970). However, these changes occur at a time in development later than the
stage corresponding to 38 in Xenopus and furthermore we have not seen any
evidence for such a reorganization in our material.
At stage 29, 30, the nuclei of the p.p. germ cells become very lobed, the nuclear
envelope appears denser and the nucleoplasm fails to take up stain so that the
nuclei as a whole have a characteristic 'empty' appearance. We suggest that the
migration of the p.p. germ cells and the change in appearance of the nuclei
are related phenomena, representing the first overt signs of p.p. germ cell
differentiation from the endoderm cells.
A careful examination of the stages immediately following 38 provides a
description of the way in which the p.p. germ cells leave the endoderm at this
time. At stage 38, 39 (there is some variation between different embryos), the
cells lie just beneath the top of the endoderm and the two sheets of lateral plate
mesoderm have not yet met in the midline (Fig. 4 A). Only slightly later, at
stage 39, 40, the p.p. germ cells have left the endoderm and lie along the endodermal crest just inside the lateral plate mesoderm, which by this time forms a
continuous sheet around the endoderm, although still only one cell thick over
the crest (Fig. 4B). As the mesoderm forms the dorsal mesentery, the p.p. germ
cells appear to be carried with it, so that at stages 40, 41, they lie embedded in
mesoderm beneath the notochord in what will become the median genital ridge
(Fig. 4C). The movements of the p.p. germ cells out of the endoderm appear to
Germ plasm and germ cells in Xenopus
A
D
Fig. 3. Diagrams to indicate the position of the presumptive primordial germ cells
in embryos of different ages: (A) morula, (B) blastula, (C) mid-gastrula, (D) completed gastrula, (E) early tail-bud (stage 20) and (F) late tail-bud (stage 29). At morula,
the germ cells are located at the vegetal pole but in the blastula they can be found
at any of the positions indicated by asterisks. During gastrulation the germ cells,
with the endoderm cells, are transported by the morphogenetic movements into the
interior of the embryo to lie in the floor of the archenteron in the completed gastrula.
At stages 20 and 29 the germ cells occupy the same relative positions in the endoderm,
although in the latter stage the archenteric cavity has changed its position as the
endoderm has elongated in the dorso-ventral aspect.
63
64
P. MCD. W H I T I N G T O N AND K. E. DIXON
Fig. 4. Stages in the migration of germ cells from the endoderm: (A) stage 38, 39;
the germ cells (gc) lie at the top of the endoderm and the two mesodermal sheets (m)
have not yet met in the midline. (B) Stage 39,40; the germ cells have left the endoderm
and lie beneath the now continuous mesodermal sheet. (C) Stage 40, 41; the germ
cells have left the endoderm and, as the dorsal mesentery (dm) forms, the germ cells
are carried and/or migrate towards the top of the mesentery where they will form
the median genital ridge.
be active, but their passage to the median genital ridge is more likely to be
partly active and partly passive. It was frequently noted that a few of the p.p.
germ cells were embedded in the dorsal mesentery about half-way along its
length, presumably carried there by the movement of the mesodermal cells.
Since, at later stages, p.p. germ cells are never seen in this position, we infer
that they are able to continue their migration within the mesoderm. These
observations and conclusions are in substantial agreement with many earlier
reports in other anuran species (see, for example, Bouin, 1900; Allen, 1907;
Witschi, 1929; Gheng, 1932).
Measurements of the position of the p.p. germ cells in the antero-posterior
Germ plasm and germ cells in Xenopus
65
Table 3. Position of germ cells in antero-posterior and dorso-ventral axes
Dorso-ventral position
Antero-posterior position
Stage
13
19,20
29-31
38-41
No. of
No. of
embryos germ cells
5
4
7
8
35
19
56
101
Distance from
anus (/tin)
(M± S.D.)
323-7±58-3
197-4 ±21 -9
316-5 ±114-8
517-3 + 429-1
Distance from
ventral edge
No. of
No. of
of endoderm
embryos germ cells (jam) (M± S.D.)
NR*
5
6
6
NR*
46
21
88
NR*
10-7 ±2-5
90±21
20-3 ±2-4
* NR = not recorded.
axis (Table 3) show that at stage 19, 20, the p.p. germ cells are still grouped
together but somewhat closer to the anus than in the gastrula. At stages 29, 30
and 38-41, the p.p. germ cells are more scattered and farther away from the
anus than at stages 19, 20. It seems likely that the overall changes in the shape
of the embryo have contributed to these changes in the positions of the p.p.
germ cells relative to the anus. Consequently, we are unable to determine whether
there is any active migration of the p.p. germ cells in the antero-posterior axis
and the measurements therefore merely record the location of the p.p. germ
cells at these stages.
Appearance of germ plasm
The appearance of the germ plasm in sections examined in the light microscope
is similar at all stages of development up to stage 38. It can be distinguished as
granular yolk-free patches within the cytoplasm (Fig. 5 A), the granularity being
most evident in 1 /im plastic sections. The patches are commonly surrounded
by numerous small, yellow-orange yolk platelets (apparent when stained with
Azan). The granularity of the germ plasm is probably due to the large number
of mitochondria which are associated with the germ plasm as reported in the
electron-microscope studies of Williams & Smith (1971) and Czolowska (1972).
At stage 38 the germ plasm is less granular than at earlier stages, and shortly
after this it loses its ability to take up the stains.
The intracellular sites of the germ plasm
At the 2-cell stage the germ plasm is visible approximately 10-15/tm from
the surface of the egg, aggregated in small patches scattered over an area extending up to several hundred microns from the vegetal pole (Figs. 5B, 6A, B).
In the 4-cell and 8-cell stages it is contained in fewer but individually larger
patches which are still located 10-15 /mi from the egg surface but now lie very
close to the cleavage membrane (Figs. 5C; 6C, D). The change in the number,
size and position of the patches of germ plasm suggests that, as the cleavage
5
E
M B
33
66
P. MCD. WHITINGTON AND K. E. DIXON
Germ plasm and germ cells in Xenopus
67
furrows form, the patches move or are moved towards the furrows, coalescing
as they go (Fig. 6E), so that about four large patches are formed within the
furrows. In embryos of 16 cells and later, approximately 80% of the p.p. germ
cells examined had a single patch of germ plasm and about 15 % had two patches
which were, in almost all cases, very unequal in size (ratios of 3:1 to 40:1). In
some embryos, however, there were numerous small patches of germ plasm (up
to 92, see also Table 4) located in a position similar to that in the freshly fertilized
egg, suggesting that aggregation and movement are causally related. Schechtman
(1934) and Ballard (1955) have described the movement of dye particles into the
interior of amphibian eggs during early cleavage. The displacement of the germ
plasm appears to be very similar to that of dye particles and therefore we conclude that the movement and aggregation of germ plasm during early cleavage
is a passive process.
The position of the germ plasm within the cleavage furrows determines its
distribution in later cleavages. Observation of p.p. germ cells in mitosis showed
that in 85 % of the cells, the germ plasm was located at one pole of the spindle
(Fig. 5C) and in only 7 of 49 mitoses was the germ plasm located in an equatorial
position, where it would have been distributed to both daughter cells. Hence
division of p.p. germ cells can continue throughout cleavage without increase
in their number.
Measurements of the position of the germ plasm relative to the plasma membrane indicate that it maintains a close association with the membrane throughout cleavage (Table 4). The apparent movement of the germ plasm towards the
plasma membrane may be due to the great reduction in size of the p.p. germ
cells, with a corresponding increase in the proportion of the cytoplasm occupied
by the germ plasm; at the 16-cell stage the germ plasm represents only 0-6 x
10~5% of the cytoplasm whereas at blastula it occupies 4 % of the cytoplasm.
Early in gastrulation the germ plasm moves to the nucleus, forming a cap or
Fig. 5. Photomicrographs of cells containing germ plasm in embryos at different
stages of development.
(A) The appearance of the germ plasm in an 8-cell stage embryo in which it forms
a relatively large, yolk-free, granular aggregate, surrounded by yolk granules which
are smaller than the large granules found in the general cytoplasm, gp, Germ plasm;
yg, small yolk granules.
(B) The position of the germ plasm in the 2-cell stage, horizontal section taken
15 /tm above the vegetal pole. The patches of germ plasm (gp) are small and numerous
in both blastomeres and arranged along the cleavage membrane.
(C) The position of the germ plasm in the 4-cell to 8-cell embryo, vertical section
showing vegetal pole (vp), the cleavage membrane (cm) and a single large patch of
germ plasm lying along the cleavage membrane. The germ plasm (gp) is at one of the
poles of the spindle and therefore when mitosis occurs it will be segregated to the
daughter cell which will lie over the vegetal pole.
(D) The position of the germ plasm in a late gastrula embryo after it has moved to its
perinuclear position, characteristically eccentrically placed and closely adherent
to the nucleus, n, nucleus; a, archenteron.
5-2
68
P. MCD. W H I T I N G T O N AND K. E. DIXON
Fig. 6. Diagrammatic representation of the position and state of aggregation of
germ plasm during early cleavage stages. (A) The 2-cell stage, vertical section, where
the germ plasm is spread over the vegetal pole and forms numerous small aggregates.
(B) View of the vegetal pole in the 2-cell stage showing the distribution of the germ
plasm beneath the vegetal surface. (C) Four-cell stage in vertical section with the
germ plasm aggregated together into one large patch situated along the cleavage
membrane. (D) Four-cell stage in horizontal section showing the similar positions of
the germ plasm aggregates in each of the blastomeres. (E) Two-cell to 4-cell stage
showing the processes of aggregation and involution of the germ plasm; as it is
moved towards the interior of the embryo along the cleavage membrane it forms a
single, large aggregate.
ring around the nucleus (Fig. 5D) as described earlier by Bounoure (1934) and
Bladder (1958). It remains in this position until stages 38-41. As a consequence
of the perinuclear position of the germ plasm, it would be distributed to both
daughter cells at mitosis and hence the number of p.p. germ cells would increase
as already noted from the direct counts of cell numbers. The movement of the
germ plasm to the nucleus may represent the first indication that the germ plasm
has become active.
Germ plasm and germ cells in Xenopus
69
Table 4. The number and position ofpatches and total volume
per embryo of germ plasm
Stage
No. of
embryos
No. of
cells
8- to 32-cell
Morula-early blastula
Mid-late blastula
9
4
9
38
20
52
Mean no. of
patches of
germ plasm
per embryo
5-2*
5t
6-9%
Vol. of germ
plasm per
embryo
(jim3 x 102)
( M ± S.D.)
Mean distance from
plasma
membrane
(/*m)
350 ±103
429 ±72
38O±198
11-2
8-7
4-1
* Two cells (not included) contained 44 and 92 patches respectively.
t Two cells (not included) contained 23 and 30 patches respectively.
t Two cells (not included) contained 16 and 29 patches respectively.
The amount of germ plasm
Quantitative estimates of the amount of germ plasm at different developmental stages show that: (i) the size of different patches within a single blastomere varies considerably (mean ratio of largest to smallest, 8:1); (ii) the amount
of germ plasm in different blastomeres varies; commonly one or two of the
blastomeres in an embryo contain most (up to 85%) of the germ plasm; (iii)
there appears to be no change in the total amount of germ plasm during cleavage
(Table 4).
The variability in the amount of germ plasm between patches, between
blastomeres and between embryos is a reflexion of the variability of the distributive processes involved and an indication of the variability that can be expected
in the final result, i.e. in the number of germ cells produced. The degree of
variability may also be some indication that the germ plasm is not active over
this period, if activity requires some measure of control over distribution. The
observation that some blastomeres contain very small quantities of germ plasm
raises the question as to whether a threshold quantity is necessary to define a
cell as a p.p. germ cell. We have never observed a cell containing germ plasm
which is not also exhibiting the behaviour of a p.p. germ cell, but nevertheless
it seems logical that a certain minimum quantity of germ plasm is required.
It has been suggested earlier that the movement of the germ plasm towards
the nucleus in the early gastrula may indicate that it is becoming 'active'. Conversely, the absence of any obvious change in the amount of germ plasm during
cleavage may lend some support to the suggestion that the germ plasm is inactive
during cleavage if it can be assumed that activity involves utilization and consumption. On these grounds also it seems unlikely that synthesis of germ plasm
takes place during this part of development. Kerr & Dixon (1974) have made
the much more likely suggestion that synthesis of germ plasm takes place during
the period of rapid mitosis in the early and developed gonad.
70
P. MCD. WHITINGTON AND K. E. DIXON
DISCUSSION
The observations reported here show that the formation of the p.p. germ cells
in the embryo is a process which can be explained in terms of normal embryological mechanisms. The germ plasm is initially partitioned between the first
four blastomeres by the first two cleavage divisions. During these cleavage
divisions it is moved by the processes associated with cleavage to a polar position
within the blastomeres where, in normal circumstances, it will be segregated to
only one of the daughter cells at subsequent divisions. Therefore, although the
four blastomeres which contain germ plasm at the 8-cell stage will divide about
10 more times up to early gastrula, as shown by reduction in their size, the
number of p.p. germ cells does not increase significantly. However, the processes
are not perfect. If, for example, the germ plasm fails to aggregate completely,
as we have occasionally observed, the chance that it will be distributed between
more cells is increased. Similarly, if the relative positions of the germ plasm and
the spindle vary from the 'normal' pattern then the chances are again greater
that the number of germ cells will increase. Our observations of the pattern of
early cleavage show that the position of the early cleavage planes is not invariable.
In any event, normal processes will eventually enhance the chances of the
number of p.p. germ cells increasing. For example, as the size of the blastomeres
decreases during cleavage, the proportion of the cytoplasm occupied by the
germ plasm increases and consequently, so does the chance that it can be incorporated into both daughter cells. But notwithstanding the inherent variability
of the embryological processes involved, we conclude that during cleavage, the
germ plasm is distributed between a small number of cells by processes which
essentially rely on simple geometric relationships between the position of the
germ plasm and the position of the cleavage planes. In the early gastrula, the
germ plasm moves to a perinuclear position, thus ensuring that as division
continues, both daughter cells will receive germ plasm and therefore the number
of p.p. germ cells will increase.
In more general terms, the processes which occur during cleavage are concerned with segregation of the germ plasm to particular blastomeres, and the
divisions which occur during later embryonic development can be interpreted
as cloning divisions which will increase the number of cells which carry germ
plasm. In this way, normal embryological processes are used to explain the
segregation of a morphogenetic determinant to particular cells in the embryo
and the subsequent first phase of proliferation of this clone of cells as they move
to their definitive site within the embryo.
The increase reported here in the number of p.p. germ cells between early
gastrula and stages 38-41 occurs at a similar time, i.e. during migration to the
genital ridges, as in mammals and birds (see Mintz, 1959). However, earlier
descriptions of p.p. germ cells in anuran amphibians suggested that division
does not take place at this time. Bounoure (1934) and later Blackler (1958)
Germ plasm and germ cells in Xenopus
71
Table 5. Numbers of germ cells recorded in anuran embryos
Means and standard deviations where possible; figures in parentheses are numbers
of observations where cited.
Stage of development
Species
X. la e vis
R. temporaria
Cleavage
Blastula
Gastrula
—
—
—
—
—
—
—
4-5 ±0-9
(13)
—
5-3± 1-9
(10
10-15
8-7 ±3-0
—
5-7
11-23
—
—
—
—
—
—
—
8-8 + 2-7f
(9)
—
Genital ridge
22-4±14-l*
(12)
29.4+11*
(12)
22-4 ±4-6*
(28)
13-9±4-8
(11)
45-9±23-3f
(10)
39 ±6-8
Reference
Nieuwkoop &
Suminski(1959)
This paper
Bounoure(1925,
1930, 1931)
Blackler (1958)$
(5)
jR. esculenta
R. pipiens
(13)
8-6±4-4f
(12)
—
—
17
—
—
—
—
—
—
—
—
—
—
B. bufo
B. viridis
D. pictus
—
—
65-7±30-9f§ Kuschakewitsch
(1910)
(7)
26
Blackler (1958)
—
DiBerardino(1961)
61, 82||
15
21-3 ±2-1
120-4± 31-8^1
(19)
24-6
Smith (1966)
Blackler (1958)
Gipouloux(1970)
Beccari (1924)
Gipouloux (1972)
* Recalculated.
t Means and standard deviations calculated from author's data.
% Neurula and hatching stages recorded as having the 'same' number as the gastrula.
§ Omitting data from tadpoles of length 9-5 mm or more following Witschi (1914).
|| Two specimens only at stage 25 (see table 3 of Smith, 1966).
II Omitting data from tadpoles of length 13 mm and greater.
claim that mitosis in the p.p. germ cells is inhibited and the cause of the inhibition is the movement of the germ plasm to its perinuclear position. Their
conclusions are based on a failure to observe any p.p. germ cells in mitosis. A
total of approximately 12 mitoses (4+8) takes place in the p.p. germ cells during
this period and therefore the chance of detecting one of these is very slight.
Furthermore, a careful examination of reports of the numbers of p.p. germ cells
at different stages of development in other anuran species negates their contention (Table 5). Two conclusions may be drawn from the data in Table 5. In
pregastrular stages of development, the number of p.p. germ cells is small and
within the range expected if geometrical processes similar to those in X. laevis
72
P. MCD. WHITINGTON AND K. E. DIXON
operate in these other species, as we would predict. The second conclusion is
that the number of p.p. germ cells in the median and early lateral genital ridges
(where mitosis does not occur) is greater than at gastrula, and 2-3 mitoses of
each p.p. germ cell present at gastrula are required to account for the increase
in number. In a group of Xenopus used in other studies in our laboratory, the
mean number of p.p. germ cells in the median genital ridge is 28 (Ziist & Dixon,
unpublished), indicating that in X. laevis there may be up to three divisions of
each p.p. germ cell, on the average, between gastrulation and stages 38-41.
It has often been claimed that the number of p.p. germ cells varies too much
between individuals and between batches of animals for any meaningful use to
be made of counts of p.p. germ cells. Although we agree that the number of p.p.
germ cells is variable both between individuals and between batches, the data
in Table 5 suggest that the variability is similar to that of many biological parameters and not so large that comparison is impossible. If anything, the data in
Table 5, compiled as they are from many different reports, emphasize the great
similarity in the number of p.p. germ cells in different genera and species at the
same developmental age, suggesting that similar processes are responsible for
p.p. germ cell segregation and proliferation in these frogs.
The question as to whether the cells do undergo mitosis or not between
gastrulation and their arrival in the genital ridges is important, because one of
the current hypotheses of the action of germ plasm is predicated on an absence
of mitotic activity at this time. Bladder (1958, 1965, 1966, 1970) has suggested
that the p.p. germ cells are able to retain their totipotency because they do not
divide, whereas the other cells of the embryo become differentiated as they
divide. This hypothesis is an attractive one, particularly because of the long
suspected but ill-defined relationship between cell differentiation and cell
division (see, for example, Holtzer, Weintraub, Mayne & Mochan, 1972). However, our results and our interpretation of other reports show that mitosis does
take place, and at approximately the same rate as in endodermal cells. From
gastrula to the late tail-bud stage, the number of endodermal cells increases
from approximately 15000 to approximately 25000 (Graham & Morgan, 1966),
a comparable if slightly lower increase to that shown by the p.p. germ cells.
The p.p. germ cells and the endoderm cells therefore have similar proliferative
patterns, which appears to invalidate the basis of Bladder's hypothesis.
We have suggested that up to the time the p.p. germ cells begin their migration
their behaviour does not differ from that of adjacent endodermal cells. We also
infer from our observations on the germ plasm that it is inactive during cleavage
but that it may become active soon after, at gastrula. Taken together, these two
suggestions imply that the cells are not overtly differentiated until they begin
to migrate, but that the process of differentiation may begin early in gastrulation,
perhaps initiated by the movement of the germ plasm to the nucleus. It is now
possible, for the first time, to investigate the initial steps in the differentiation
of a particular cell line and work is proceeding in our laboratory on this problem.
Germ plasm and germ cells in Xenopus
73
Nevertheless, the role of the germ plasm in defining the p.p. germ cells remains
obscure. Kerr & Dixon (1974) have suggested that the germ plasm may play a
part in the following processes: (i) the migration of the presumptive primordial
germ cells from the endoderm to the genital ridge, (ii) mitosis of the primordial
germ cells and subsequently of the oogonia and spermatogonia in the developing
gonads, and (iii) preparations for meiosis. In suggesting that the germ plasm is
inactive during early embryogenesis, these results lend some support to these
suggestions.
Our general conclusions from this study therefore are that the germ plasm is
segregated to a small number of blastomeres during cleavage. Differentiation of
the presumptive primordial germ cells (as they can then more properly be called)
and cloning divisions follow their segregation as would be expected from general
embryological theory.
Our thanks are due to Dr P. A. Janssens, The Australian National University, and Mr
H. A. Clair, the University of Western Australia, for help in providing additional Xenopus,
and to Miss B. Ziist for advice and assistance. Mr G. Sobels and Mrs E. Harrland gave
valuable technical assistance.
REFERENCES
B. M. (1907). An important period in the history of the sex cells of Rana pipiens.
Anat. Am. 31, 339-347.
BALLARD, W. W. (1955). Cortical ingression during cleavage of amphibian eggs, studied by
means of vital dyes. /. exp. Zool. 129, 77-98.
BALLARD, W. W. (1970). Archenteric origin of midgut lumen in Amphibia. Devi Biol. 21,
424-439.
BECCARI, N. (1924). Studi sulla prima origine delle cellule genitali nei vertebrati. III. Richerche
nel Bufo viridis. Archo Hal. Anat. Embriol. 21, 332-374.
BLACKLER, A. W. (1958). Contribution to the study of germ cells in the Anura. /. Embryo/.
exp. Morph. 6, 491-503.
BLACKLER, A. W. (1965). The continuity of the germ-line in amphibians and mammals.
Annee Biol. 4, 627-635.
BLACKLER, A. W. (1966). Embryonic sex cells of Amphibia. Adv. Reprod. Physiol. 1, 9-28.
BLACKLER, A. W. (1970). The integrity of the reproductive cell line in the Amphibia. Curr.
Top. Devi Biol. 5, 71-87.
BOUIN, M. (1900). Histogenese de la glande genitale femelle chez Rana temporaria L. Archs
Biol., Paris 17,201-381.
BOUNOURE, L. (1925). L'origine des gonocytes et revolution de la premiere ebauche genitale
chez les Batraciens. Annls Sci. not. 10e ser., 8, 201-278.
BOUNOURE, L. (1930). Sur la presence de cellules germinales distinctes dans la blastula de la
grenouille rousse. C. r. hebd. Seanc. Acad. Sci., Paris 190, 1143-1145.
BOUNOURE, L. (1931). A quel moment peut-on parler de cellules germinales initiates dans le
developpement de la grenouille rousse? C. r. hebd. Seanc. Acad. Sci., Paris 107, 988-991.
BOUNOURE, L. (1934). Recherches sur la lignee germinate chez la grenouille rousse aux premiers
stades du developpement. Annls Sci. nat. 10e ser., 17, 67-248.
BOUNOURE, L. (1964). La lignee germinate chez les batraciens anoures. In L'origine de la lignee
germinate (ed. E. Wolff), pp. 207-234. Paris: Hermann.
BUEKR, M. L. & BLACKLER, A. W. (1970). Sterility and partial sterility in the South African
clawed toad following the pricking of the egg. /. Embryol. exp. Morph. 23, 375-384.
CHENG, T. H. (1932). The germ cell history of Rana cantabrigensis Baird. 1. Germ cell origin
and gonad formation. Z. Zellforsch. mikrosk. Anat. 16, 497-541.
ALLEN,
74
P. MCD. W H I T I N G T O N AND K. E. DIXON
R. (1972). Thefinestructure of the 'germinal cytoplasm' in the egg of Xenopus
laevis. Wilhelm Roux Arch. EntwMech. Org. 169, 335-344.
Di BERARDINO, M. (1961). Investigations of the germ plasm in relation to nuclear transplantation. /. Embryol. exp. Morph. 9, 507-513.
GIPOULOUX, J. D. (1970). Recherches experimental sur l'origine, la migration des cellules
germinales, et, l'edification des cretes genitales chez les Amphibiens Anoures. Bull. biol.
Fr. Belg. 104, 21-93.
GIPOULOUX, J. D. (1972). Regulation du nombre des cellules germinales apres un exces
initial de celles-ci dans la gonade larvaire des Amphibiens Anoures. C. r. hebd. Seanc. Acad.
ScL, Paris 274, 2229-2231.
GRAHAM, C. F. & MORGAN, R. W. (1966). Changes in the cell cycle during early amphibian
development. Devi Biol. 14, 439-460.
GURR, E. (1962). Staining Animal Tissues, Practical and Theoretical. London: Hill.
HARRIS, T. M. (1967). The morphogenesis of the stomach and intestine in the salamander
Ambystoma maculatum. J. Morph. 122, 345-366.
HOLTZER, H., WEINTRAUB, H., MAYNE, R. & MOCHAN, B. (1972). The cell cycle, cell lineages,
and cell differentiation. Curr. Top. Devi Biol. 7, 229-256.
KERR, J. B. &DIXON, K. E. (1974). An ultrastructural study of germ plasm in spermatogenesis
of Xenopus laevis. J. Embryol. exp. Morph. 32, 573-592.
KUSCHAKEWITSCH, S. (1910). Die Entwicklungsgeschichte der Keimdrusen von Rana
escu/enta. In Festschr. sechzign. Geburts. H. Hertwig, 2, 61-224. Jena: G. Fischer.
MINTZ, B. (1959). Continuity of the female germ cell line from embryo to adult. Archs Anat.
microsc. Morph. exp. 48 (suppl.), 155-172.
MCCLUNG, C. E. (1964). Handbook of Microscopical Technique, 3rd ed. New York: Hafner.
NIEUWKOOP, P. D. & SUMINSKI, E. H. (1959). Does the so-called 'germinal cytoplasm' play
an important role in the development of the primordial germ cells. Archs Anat. microsc.
Morph. exp. 48 (suppl.), 189-198.
NIEUWKOOP, P. D. & FABER, J. (1967). Normal Table of Xenopus laevis (Daudin). 2nd ed.
Amsterdam: North-Holland Publ. Co.
SCHECHTMAN, A. M. (1934). Unipolar ingression in Triturus torosus: a hitherto undescribed
movement in the pregastrular stages of a Urodele. Univ. Calif. Publs Zool. 39, 303-310.
SMITH, L. D. (1966). The role of a 'germinal plasm' in the formation of primordial germ
cells in Rana pipiens. Devi Biol. 14, 330-347.
TAHARA, Y. & NAKAMURA, O. (1961). Topography of the presumptive rudiments in the
endoderm of the Anuran neurula. /. Embryol. exp. Morph. 9, 138-158.
VOLKONSKY, M. (1928). Sur une nouvelle modification de la technique d'Altmann. Bull.
Histol. appl. Physiol. Path. 5, 220-222.
WILLIAMS, M. A. & SMITH, L. D. (1971). Ultrastructure of the 'germinal plasm' during
maturation and early cleavage in Rana pipiens. Devi Biol. 25, 568-580.
WEISMANN, A. (1885). Die Continuitdt des Keimplasmas als Grundlage einer Theorie der
Vererbung. Jena: Fischer.
WITSCHI, E. (1914). Experimentelle Untersuchungen iiber die Entwicklungsgeschichte der
Keimdrusen von Rana temporaria. Arch, mikrosk. Anat. EntwMech. 85, 9-113.
WITSCHI, E. (1929). Studies on sex differentiation and sex determination in amphibians.
/. exp. Zool. 52, 235-265.
WOLF, D. P. & HEDRICK, J. L. (1971). A molecular approach to fertilization. II. Viability and
artificial fertilization of Xenopus laevis gametes. Devi Biol. 25, 348-359.
CZOLOWSKA,
{Received 17 April 1974)