Download Embryological evidence for a possible

/. Embryo!, exp. Morph. Vol. 35, 1, pp. 159-167, 1976
159
Printed in Great Britain
Embryological evidence for a possible polyphyletic
origin of the recent amphibians1
By P. D. NIEUWKOOP 2 AND LIEN A. SUTASURYA 3
From the Hubrecht Laboratory, Utrecht, and the
Department of Biology, Bandung
SUMMARY
The markedly different mode of mesoderm formation in anuran and urodelan amphibians
(which is related to the early double-layered nature of the anuran blastula wall in contrast
to its single-layered nature in the urodeles), but particularly the fundamentally different
place and mode of origin of the primordial germ cells in the two groups of amphibians,
strongly pleads in favour of a very ancient bifurcation in the phylogenetic history of the two
groups, even suggesting a polyphyletic origin from different ancestral fishes.
INTRODUCTION
Although Haeckel's biogenetic law (1891, 1894) - according to which
animals pass through the same stages in their embryonic development as they
did in their phylogenetic history - is no longer considered as having general
validity, it cannot be denied that related species show a nearly identical embryonic development up to a relatively advanced stage, after which their development begins to diverge. This is the consequence of the fact that related species
have a large number of genes in common. Animals which are not so closely
related generally diverge in their embryonic development at an earlier stage.
Finally, the development of species which are hardly related at all differs from
the very beginning or shows a very short common phase only, after which each
follows its own separate course. (It should, however, be realized that these
considerations cannot simply be generalized, since at any stage of development
the embryo represents a functional unit which vitally depends upon its environment. Unrelated forms may therefore be subject to similar environmental
conditions leading to common features in their development.) In our opinion,
the converse is therefore also true: the earlier and the more pronounced the
differences in embryonic development of the two species or two animal groups,
the wider they are apart genetically. Accepting the theory of evolution for the
1
Dedicated in personal friendship to Prof. Dr Etienne Wolff on the occasion of his
retirement.
2
Author's address: Hubrecht Laboratory, Utrecht, Holland.
3
Author's address: Department of Biology, I.T.B., Bandung, Indonesia.
160
P. D. NIEUWKOOP AND L. A. SUTASURYA
explanation of the diversity of forms in the animal and plant kingdoms, this
would mean that forms which differ strongly in development must have developed along different evolutionary pathways from an early phase of their
phylogenetic history. We would like to review some features in the development
of the two groups of the living infraclasses of amphibians, namely the anurans
(belonging to the Salientia) and the urodeles (belonging to the Caudata) in the
light of the above considerations. We must leave the rather unknown order of
the Caecilians out of consideration for lack of relevant information.
OBSERVATIONS: DISCUSSION
Comparison of the embryonic development of anuran and
urodelan amphibians
It is evident that the anuran and urodelan amphibians differ anatomically in
many respects. This becomes particularly evident during metamorphosis, when
the anuran tadpole changes into a tailless frog or toad, whereas the urodelan
larva hardly changes in general appearance. It is, however, not our intention to
discuss here the anatomical features of representatives of the two groups, but
to focus attention upon their early development.
When comparing the maps of presumptive organ rudiments in early embryos
of the two groups (see the anlage maps of Triturus and Bombina, Vogt (1929))
there is, among other things, a striking difference in the width of the marginal
zone. At the early blastula stage {ca. stage 8 Harrison) the marginal zone
comprises a sector of about 60° on the dorsal side in Triturus (cf. Fig. 1 d),
but of only 45° in Bombina (cf. Fig. \c). The presumptive neural area is also
markedly larger in Triturus than in Bombina: 65° as against 35°. Whereas the
future anterior end of the brain coincides with the original animal pole of the
Triturus egg, it lies 35° away from the animal pole in the Bombina blastula;
consequently in the Bombina larva the original animal pole corresponds to a
point somewhere in the cranio-ventral epidermis of the heart region. These
differences, however, only concern the relative dimensions of the organ anlagen.
Although they are quite pronounced, it is not self-evident that they are of a
fundamental nature.
A much more fundamental difference between anuran and urodelan development concerns the position of the marginal zone, i.e. the presumptive endomesodermal anlage, in the two groups. In the anuran blastula/early gastrula
the marginal zone is exclusively internally located (cf. Fig. 1 c), in contrast to
its external position in the urodelan blastula (cf. Fig. 1 d). This difference was
not clearly understood by Vogt (1929), but was extensively described by
Nieuwkoop and Florschutz (1950) and recently verified by Keller (1975) by
means of careful vital staining experiments. The internal position of the presumptive chordal and prechordal mesoderm in the anuran blastula (see also
Pasteels (1949) for the position of the prechordal endo-mesoderm) is clearly
Polyphyletic origin of recent amphibians
A.P.
A.P.
v.p.
V.P.
161
Fig. 1. Diagrammatic representation of mesoderm formation and origin of PGCs
in the anuran and urodelan amphibians, (a) Location of the germinal cytoplasm in
the vicinity of the vegetative pole of the anuran egg. (b) Absence of germinal cytoplasm in the urodelan egg. (c) Location of the PGCs in the floor of the blastocoel,
and formation of the relatively narrow, exclusively internal marginal zone in the
anuran blastula. (d) Formation of the more extensive and externally situated
marginal zone, and origin of the PGCs from the latero-ventral mesoderm in the
urodelan blastula. (e) Migration of the endodermal PGCs towards the future
gonadal anlagen in the young anuran larva. (/) Migration of the mesodermal
PGCs towards the future gonadal anlagen in the young urodelan larva.
II
EMB
35
162
P. D. NIEUWKOOP AND L. A. SUTASURYA
related to the double-layered nature of the blastocoelic wall, in contrast to the
essentially single-layered nature of the urodelan blastula wall. In the anuran
blastula the outer, unicellular layer gives rise to the ectodermal epithelial layer
of the epidermis and the ependymal layer of the central nervous system, as well
as to the endodermal inner lining of the archenteron, while the inner, pluricellular layer comprises the presumptive sensorial layer of the epidermis, the
bulk of the future nervous system, the entire chordal and prechordal mesoderm,
and the bulk of the endoderm (cf. Fig. \c). In the urodelan blastula there is no
sharp delimination between an outer and an inner layer and both the presumptive nervous system and the mesoderm are represented in the outer surface
(cf. Fig. Id). A segregation of the ectoneuroderm into an epithelial and a
sensorial epidermal layer and into the corresponding layers of the c.N.s. only
occurs much later in the urodelan embryo. This developmental difference
between the two groups is quite fundamental. In the anuran embryo the doublelayered character arises very early in development, namely between the 32-cell
stage (stage 6 N. and F.), which is still single-layered, and the early blastula
(stage 7 N. and F.), which is already essentially double-layered due to radial
cell divisions starting in the equatorial region (stage 6% N. and F.) and spreading
rapidly over the animal hemisphere (Nieuwkoop & Faber, 1967). A further
thinning of the outer, epithelial layer by local radial divisions continues, up to
at least the late blastula stage. The double-layered character of the anuran
embryo certainly represents a very early divergence in the development of the
two groups.
We now come to a still more fundamental difference between the two groups
of amphibians, namely the origin and development of the primordial germ cells
(PGCs).
In the anuran Rcma temporaria the development of the PGCs was studied in
detail by Bounoure in the nineteen-thirties (see Bounoure, 1939). He found
that the PGCs are characterized by a special cytoplasmic component, the
so-called 'germinal cytoplasm', which according to Mahowald & Hensen
(197J), Williams & Smith (1971) and Czolowska (1972) consists of an accumulation of mitochondria, 'germinal granules' with a particular ultrastructure,
and polyribosomes. This germinal cytoplasm can be traced back to the fertilized
egg, where it is situated under the plasmalemma in the vicinity of the vegetative
pole in the form of small islands (cf. Fig. 1 a). It could even be found in a more
fragmentary form in the mature oocyte (Czolowska, 1969). As a consequence
of the original location of the germinal cytoplasm in the vicinity of the vegetative
pole the PGCs are of endodermal origin in the anurans. According to Whitington & Dixon (1975), in Xenopus the germinal cytoplasm during cleavage
becomes distributed over a restricted number of blastomeres and often passes
into only one of the daughter cells of a dividing blastomere. They could show
that only those blastomeres which receive (sufficient?) germinal cytoplasm will
develop into PGCs. During cleavage and blastula formation they move towards
Polyphyletic origin of recent amphibians
163
the vicinity of the blastocoel (cf. Fig. 1 c) and are subsequently found in the
caudal portion of the invaginated yolk mass, from where they migrate - probably by amoeboid movement - towards the mid-dorsal endoderm, from there
into the dorsal mesentery, and finally into the germ ridges (cf. Fig. 1 e). During
neurulation the germinal cytoplasm becomes displaced within the blastomeres
from an originally peripheral to a juxtanuclear position. Blackler (1970) could
show that the PGCs represent the only cellular elements which give rise to the
next generation. In the anurans, therefore, the continuity of the germ line
throughout the entire life of the individual could be clearly demonstrated; the
PGCs being set apart as specific elements from the somatic cells at a very early
stage of development and being characterized by the presence of germinal
cytoplasm.
When we now turn our attention towards the urodeles, Humphrey (1925,
1927, 1928, 1929) already demonstrated that in an advanced tail-bud stage of
Ambystoma the PGCs are situated in the lateral mesoderm in the vicinity of the
Wolflfian ducts, and move from there towards the genital ridges after the formation of the coelomic cavities (cf. Fig. 1/). By means of extirpation and transplantation of portions of the endoderm or of the entire endoderm at gastrula to
neurula stages, Nieuwkoop (1947) could demonstrate that in the urodeles the
PGCs do not originate from the endoderm (cf. Fig. \b). Removal of the
presumptive lateral plate mesoderm at an early neurula stage led to completely
sterile larvae (Nieuwkoop, 1947). In hetero- and xenoplastic chimaerae made
by transplantation of the latero-ventral marginal zone at a mid-gastrula stage,
PGCs of both species were found in the genital ridges, and there was a clear
correlation between the regional chimaeric composition of the lateral plates and
the regional distribution of PGCs of the two species. In our opinion, these facts
irrefutably prove the mesodermal origin of the PGCs in the urodeles (cf. Fig.
Id). Smith (1964), using eggs of genetically different white and black axolotls,
could show that the PGCs which originate from the latero-ventral mesoderm
actually give rise to the next generation. In contrast to these observations the
Japanese authors Amanuma (1957), Asayama (1950, 1961) and Asayama &
Amanuma (1957) came to a different conclusion on the basis of extirpations of
lateral mesoderm. However, Capuron (1972) could demonstrate that their
extirpated material did not correspond topographically with the actual position
of the PGCs at the stages of operation. Kotani (1957) was the first to claim that
in the urodeles the PGCs originate from the ectoderm. This observation was
recently corroborated by induction experiments made by Kocher-Becker and
Tiedemann (1971). Very recently Miss Sutasurya could demonstrate that the
PGCs in the urodeles exclusively originate from the animal, ectodermal moiety
of the blastula. They are not formed from specific cellular elements set aside at
the beginning of development, but actually originate from common, totipotent
ectodermal cells under an inductive influence exerted by the ventral endodermal
yolk mass. They apparently constitute one of the characteristic cellular elements
164
P. D. NIEUWKOOP AND L. A. SUTASURYA
of the induced ventro-caudal mesoderm, just as, for example, the pigment cells
represent a characteristic element of the neural crest; for the PGCs can arise
from any portion of the animal, ectodermal moiety of the blastula. However, in
the blastula the competence for PGC formation (like that for mesoderm
induction in general) decreases from the equator towards the animal pole. In
normal development PGCs are only formed in the ventral and latero-ventral
marginal zone (cf. Fig. 1 d). This is because their formation elsewhere is suppressed by the dorsal type of mesoderm induction leading to notochord and
somite differentiation in the dorsal and lateral marginal zone (Sutasurya &
Nieuwkoop, 1974). As soon as the mesodermal competence is lost, PGCs can
no longer be induced. Smith's (1964) negative results after transplantation of
early gastrula ectoderm into the latero-ventral marginal zone are probably due
to the fact that the inducing capacity of the endoderm had faded out by the
time the operations were carried out (the mid-gastrula stage) (See Boterenbrood
& Nieuwkoop, 1973).
From this it must be concluded that some very fundamental differences in
PGC formation exist between anuran and urodelan amphibians. The PGCs not
only have an entirely different place, but also an entirely different mode of
origin in the two groups. On the other hand, it must be realized that in the
anurans, where the germinal cytoplasm can already be found in the mature
oocyte, the PGCs may nevertheless not fully segregate from somatic cell
material before an advanced gastrula stage, while in the urodeles the induction
of the latero-ventral mesoderm is not accomplished before a mid-gastrula stage.
Several important questions arise at this stage. When are the PGCs actually
formed in the urodeles ? Although in the urodelan larva the PGCs are histologically quite similar to those of the anuran larva, it is not known either whether
in the former the PGCs are also characterized by germinal cytoplasm. If so,
when and how does this arise? If not, what characterizes the PGCs in the
urodeles, at what stage of development are they actually formed, and when do
they acquire their characteristic features ?
The phylogenetic origin of the anuran and urodelan amphibians
The pronounced differences in embryonic development of the anuran and
urodelan amphibians, exemplified among other things by the differences in
mesoderm formation, but above all by the fundamental differences in PGC
formation in the two groups, in our opinion strongly suggests that the two
groups of amphibians are actually only remotely related. These fundamental
differences can only be understood if the bifurcation in the evolution of the two
groups is placed as far back as possible in the phylogenetic history of the
vertebrates.
Colbert in his monograph (1955) places the origin of both anurans and
urodeles back in the Carboniferous period, possibly from different ancestors.
In the second (1969) edition, however, he suggests on comparative-anatomical
Polyphyletic origin of recent amphibians
165
grounds that the two infraclasses Salientia and Caudata belong to the same
subclass of the Lissamphibia and have a common origin. We must say that the
arguments for the latter statement are not very convincing. He actually mentions
that not only is the origin of the anurans not yet known - the first fossil records
of the anuran ancestors being found in the lower Trias, namely the genus
Triadobatraclms from Madagascar - but the origin of the urodeles is still more
uncertain - the first urodelan fossils being known only from Cretaceous
sediments. He further suggests that the teleosts have developed in the Devonian
and Carboniferous periods from holostean and chondrostean ancestors, which
in their turn have probably originated from the first, jawless vertebrates in the
Ordovician period. The ancestors of the reptilians go back in history to the
early Carboniferous vertebrates from which the amphibians may also have
originated.
Jn her recent book Barbara J. Stahl (1974) extensively discusses the still
controversial opinions on the origin of the modern Amphibia. Whereas, for
example, Jarvik and Schmalhausen strongly defend a diphyletic origin of the
recent Amphibia, Parsons & Williams and Thomson support a monophyletic
origin; the sharply opposed opinions being mainly due to the scarcity of palaeontological evidence. It is, however, striking that in the extensive literature on
the phylogeny of the vertebrates embryological evidence is hardly taken into
consideration.
When one realizes that in the teleosts the PGCs are of endodermal origin and
are at later stages characterized by germinal cytoplasm (see Satoh, 1974, in
Oryzias) numerous questions arise. Do the teleosts actually show a similar
formation of the PGCs as the anuran amphibians ? Do the chondrosteans and
the holosteans actually have the same type of PGC formation as the teleosts?
Are there perhaps other fishes, e.g. the cartilaginous Chondrichthyes or the
jawless Agnatha (modern Cyclostomata), which show a urodelan type of PGC
formation? Are the two types of PGC formation found in anuran and urodelan
amphibians also found in different subclasses of the reptilians? etc.
Assuming that the origin of the PGCs represents a very fundamental character in vertebrate development, a further analysis of the origin of the PGCs in
the various groups may yield some further insight into the still mysterious
phylogeny of the vertebrates. Our present, though limited knowledge of PGC
formation in the two groups of amphibians in our opinion points towards a
polyphyletic origin of the two groups from different ancestral fishes rather than
towards an ancient, but common origin of the tetrapods.
166
P. D. NIEUWKOOP AND L. A. SUTASURYA
REFERENCES
A. (1957). Effect of extirpation of the presumptive intermediate mesoderm
upon the differentiation of the primordial germ cell. Zool. Mag., Tokyo 66, 310-313.
ASAYAMA, S. (1950). The developmental potencies of the intermediate mesoderm of Triturus
pyrrhogaster when transplanted into orthotopic or heterotopic site in the body wall of
another embryo. /. Inst. Polytech. Osaka City Univ. D 1, 13-28.
ASAYAMA, S. (1961). Potency of lateral plate mesoderm relating to the formation of reproductive tissue. Zool. Mag., Tokyo 70, 289-293.
ASAYAMA, S. & AMANUMA, A. (1957). On the primordial germ cells of the secondary embryo
induced by the organizer. Zool. Mag., Tokyo 66, 279-283.
BLACKLER, A. W. (1970). The integrity of the reproductive cell line in the Amphibians. Curr.
Topics Devi Biol. 5, 71-87.
AMANUMA,
BOTERENBROOD, E. C. & NIEUWKOOP, P. D. (1973). The formation of the mesoderm in
urodelean amphibians. V. Its regional induction by the endoderm. Wilhelm Roux Arch.
EntwMech. Org. 173, 319-332.
BOUNOURE, L. (1939). UOrigine des cellules reproductrices et leprobleme de la ligne germinale.
Paris: Gauthiers-Villars.
CAPURON, A. (1972). Mise en evidence de gonocytes dans divers territoires isoles au stade
du bourgeon caudal et cultives in vitro chez Pleurodeles waltlii Michah (Amphibien
Urodele). C. r. hebd. Seam. Acad. Sci., Paris D 274, 277-279.
COLBERT, E. H. (ed.) (1955, 1969). Evolution of the Vertebrates, a History of the Backboned
Animals in Time. New York: J. Wiley & Sons.
CzctOWSKA, R. (1969). Observations on the origin of the 'germinal cytoplasm' in Xenopus
laevis. J. Embryol. exp. Morph. 22, 229-251.
CzctOWSKA, R. (1972). The fine structure of the 'germinal cytoplasm' in the egg of Xenopus
laevis. Wilhelm Roux Arch. EntwMech. Org. 169, 335-344.
HAECKEL, E. (1891). Anthropogenie. Leipzig: W. Engelmann.
HAECKEL, E. (1894). Systematische Phylogenie. Berlin: Georg Reimer.
HUMPHREY, R. R. (1925). The primordial germ cells of Hemidactylium and other amphibia.
/. Morph. Physio/. 41, 1-43.
HUMPHREY, R. R. (1927). Extirpation of the primordial germ cells of Amblystoma; its
effect upon the development of the gonad. /. exp. Zool. 49, 363-399.
HUMPHREY, R. R. (1928). The developmental potencies of the intermediate mesoderm of
Amblystoma when transplanted into ventro-lateral sites in other embryos: the primordial
germ cells of such grafts and their role in the development of a gonad. Anat. Rec. 40,
67-90.
HUMPHREY, R. R. (1929). The early position of the primordial germ cells in Urodeles;
evidence from experimental studies. Anat. Rec. 42, 301-314.
KELLER, R. E. (1975). Vital dye mapping of the gastrula and neurula of Xenopus laevis.
I. Prospective areas and morphogenetic movements of the superficial layer. Devi Biol. 42,
222-241.
KOCHER-BECKER, U. & TIEDEMANN, H. (1971). Induction of mesodermal and endodermal
structures and primordial germ cells in Triturus ectoderm by a vegetalizing factor from
chick embryos. Nature, Lond. 233, 65-66.
KOTANI, M. (1957). On the formation of the primordial germ cells from the presumptive
ectoderm of Triturus gastrulae. / . Inst. Polytech. Osaka City Univ. D 8, 145-158.
MAHOWALD, A. P. & HENSEN, S. (1971). Ultrastructure of the 'germ plasm' in eggs and
embryos of Rana pipiens. Devi Biol. 24, 37-53.
NIEUWKOOP, P. D. (1947). Experimental investigations on the origin and determination of
the germ cells, and on the development of the lateral plates and germ ridges in Urodeles.
Archs. neerl. Zool. 8, 1-205.
NIEUWKOOP, P. D. &FABER, J. (1967). Normal Table o/Xenopus laevis (Daudin): a Systematical and Chronological Survey of the Development from the Fertilized Egg Till the End of
Metamorphosis. Amsterdam: North-Holland Publ. Company.
Polyphyletic origin of recent amphibians
167
P. D. & FLORSCHUTZ, P. A. (1950). Quelques caracteres speciaux de la gastrulatiori et de la neurulation de Poeuf de Xenopus laevis Daud. et de quelques autres Anoures.
A rchs Biol. Paris 61,113-150.
PASTEELS, J. J. (1949). Observations sur la localisation de la plaque prechordale et de
l'entoblaste presomptifs au cours de la gastrulation chez Xenopus laevis. Archs Biol.,
Paris 60, 235-250.
SATOH, N. (1974). An ultrastructural study of sex differentiation in the teleost Oryzias latipes.
J. Embryo I. exp. Morph. 32, 195-215.
SMITH, L. D. (1964). A test of the capacity of presumptive somatic cells to transform into
primordial germ cells in the Mexican axolotl. /. exp. Zool. 156, 229-242.
STAHL, B. j . (1974). Vertebrate History: Problems in Evolution. New York: McGraw-Hill.
SUTASURYA, L. A. & NIEUWKOOP, P. D. (1974). The induction of the primordial germ cells
in the Urodeles. Wilhelm Roux Arch. EntwMech. Org. 175, 199-220.
VOGT, W. (1929). Gestaltungsanalyse am Amphibienkeim mit ortlicher Vitalfarbung.
IF. Gastrulation und Mesodermbildung bei Urodelen und Anuren. Wilhelm Roux Arch.
EntwMech. Org. 120, 384-706.
WHITINGTON, P. MCD. & DIXON, K. E. (1975). Quantitative studies of germ plasm and
germ cells during early embryogenesis of Xenopus laevis. J. Embryol. exp. Morph. 33,
57-74.
WILLIAMS, M. A. & SMITH, L. D. (1971). Ultrastructure of the 'germinal plasm' during
maturation and early cleavage in R. pipiens. Devi Biol. 25, 568-580.
NIEUWKOOP,
{Received 5 September 1975)