INTRODUCTION
In his isolation experiments Holtfreter (1938) has systematically investigated the developmental capacities of the three presumptive germ layers of the early urodelean and anuran gastrula, demonstrating the existence of pronounced differences among them. He found that the presumptive endoderm exhibits strong developmental capacities for various endodermal structures; the presumptive mesoderm shows a broad variety of developmental capacities for mesodermal, ecto-neurodermal as well as endodermal differentiation, while the presumptive ecto-neurodermal region shows hardly any developmental capacity. The latter forms only an atypical mass of ectodermal cells.
Focussing our attention on the development of the mesoderm, Yamada (1937, 1940) demonstrated that the medio-lateral organization of the mesoderm of Triturus pyrrhogaster depends upon an inductive action emanating from the medio-dorsal noto-chordal anlage and which spreads with decrement laterally and ventrally.
Toivonen (1953) showed that mesodermal structures of Triturus vulgaris can be induced in the gastrula ectoderm by using guineapig bone marrow as heterogenous inductor. Yamada (1958) found that bone marrow extract induced not only mesodermal but also endodermal structures in the gastrula ectoderm. This was later confirmed by several authors, e.g. Takata and Yamada (1960), Masui (1961), Ogi (1961), Engländer (1962) and Tseng (1963).
Studying the loss of mesodermal as well as neural competence of the presumptive ecto-neuroderm of Triturus vulgaris during gastrulation by using heterogenous inductors, Leikola (1963) has come to the conclusion that there are two kinds of competence: neural and mesodermal, the latter being lost earlier than the former.
Gebhardt and Nieuwkoop (1964), studying the mesodermizing action of lithium ions in axolotl, demonstrated that the blastula ectoderm has a much higher mesodermization competence than the gastrula ectoderm. The latter loses its competence at about the horse-shoe-shaped blastopore stage (stage \(10\frac{1}{2}\)),
Harrison). The action of lithium ions was shown to be accompanied by extensive cytolysis of the ectodermal cells.
Also using lithiuur j-ons as inductor, Grunz (1968) investigated the endo- and mesodermal competence of the presumptive ectoderm at successive stages of development in two urodeles. He demonstrated Ehat competence begins to appear at the morula stage, reaches a maximum at the middle to late blastula and termj.nates at the early gastrula stage. Whereas the ectodern of Atnbystoma forrns endo- and mesodermal structures of tai-1 character, that of Tr.LtunLs ctriefly produces th-ose characteristic of the posterior and middle trunk region.
In his recombination experiments with micromeres and macromeres encircling the animal and vegetative poles of the rnorula to blastula stages of Ttitutas pyrrhogaster, Ogi (7967) observed that although the tlro parls did not contain any presumptive rnesodermal cells, mesoderm formation occurred in the recombl-nates. He inferred the presence of a double gradient system in amphibian development analogous to that demonstrated in lhe sea urchin (Hiirstadius, 1962; Runnstrijm, 1966) .
The present study was carried out to gain information on mesodermal competence of the presurnptive ecto-neuroderrn with further development, in Xenopus Lqeuis. For this purpose experj-ments were designed to observe the reactive capacity of the presumptive ecto-neuroderm at various developmental stages.
MATERIALS AND METHODS
Throughout the investigation, eggs of Xenopus La.euis (South African clawed frog) reared at the Hubrecht Laboratory, Utrecht-Holland, were used. The eggs were obtained by injecting the males and females with Physex (Leo Pharrnaceutical Products, Ballerup-Denrnark), a gonadotrophin. The injection was made into the dorsal lymph sac, piercing the skin of the thigh and the septum between the lyrnph sac of the thigh and the back. Two days before the eggs were required, the males were injected with 60 I.U. in 0.5 cc of aquadest. On the next afternoon they r^rere once again injected with the same dose, and the females with 200 I.U. in 1 cc of aquadest. The animals were put in a covered container, sheltered from 1ight, and provided w1th plastic bars for ovipositlon. Spawning occurred in the early morning, after the temperature had been raised artificially during the night and had passed the critical temperature which ls 21"C.
To test the reactive capacity of the presumptive ectoneuroderm before and during gastrulation to the inductive influence emanating from the endoderm, different ages of the presumptive ecto-neuroder:rn ranging from early bl-astula to lrorse-shoe-shaped blastopore stage (stage 6rr/7 to stage 11) were combi-ned with endoderm of late blastula (stage 9l Nieuwkoop and Faber L967). Both parts of the recombinates hTere reared as isolates for controls. The recombinates as well as the j-solates were cultured for three days at 20oC in normal modified Barthts solution, to which 20.000 I.U./L penicillin and 0.I g/f streptomycin were added as antibacterial agents.
A11 specimens were fixed in Srnithrs solution for six hours, then washed directly in 701l alcohol-. They were then block-stained with borax-carmine and subsequently ernbedded in paraffin-wax through n-butyJ- alcohol and -ctioned at 10/U. The sections were counterstained with aniline blue-orange G.
RESULTS
The isolates of the presumptive ecto-neuroderrn (zone I-II) formed irregul-ar masses of atypical ectoderm, whereas the isolates of presumptive endoderm (zone IV) each consisted of a solid mass of undifferentiated yolk-laden cel1s.
Recombinates were obtained by placing zone IV into the cup-shaped zone I-II (Fig. 1). The two parts soon adhered to each other firmly and the recombinates hrere then turned to their normal position. Development of all recombinates was mainly characterized by exogastrulation. The older the gastrula, the smaller became the area of free presumptive ectoderm. The cornbining of later stages of presumptive ectoderm with the endoderm of late blastula did not show any rejection phenomena.
In recombinates, ectodermal. derivatives formed were epidermis, placodes, melanophores and neural tissues, the latter being present in the form of either a distinct neural tube or an irregular neural mass. The mesodermal structures consisted of muscle tissue, notochord, nephric tubules, blood ce11s and mesenchyme. Most of the endoderm remai-ned undifferentiated and was found as solid mass of yolk-laden cel-1s.
The experi.mental results are sunmarized in table 1. In all experimental series the percentage of cases with neural differentiation lies between that \,rith notochordal and muscle differentiation. Using presumptive ectoderm as reactive tissue in pre-gastrulation stages, muscle was found in L007", while notochord in 40 - 451l of the cases. The frequency of occurrence of muscle and notochord differentiation dropped very rapidly in gastrulation stage. In recombinates with presumptive ectoderm of the horse-shoe-shaped blastopore stage (stage 11), muscle differentiation was found in only 25% and notochord in 5% of the cases. Blood cells started to differentiate in recombinates with presumptive ecto-neuroderm of the l "
Fig. 1. Diagrammatic representation of the operation, showing the areas taken from the blastula to form the recombinate of the entire presumptive ecto-neuroderm (zone I \(\sim\) II) with the presumptive endoderm (zone IV).
early blastula stage (stage 7<sup>-</sup>) and was formed in 5% of the cases. In recombinates with older presumptive ecto-neuroderm the frequency of occurrence increased rapidly, especially that of the crescent-shaped blastopore stage.
At a later stage, the horse-shoe-shaped blastopore or stage 11, there was a pronounced drop. Yet, the frequency of blood cell occurrence in the latter case was still higher (35% of the cases) than that of other main mesodermal structures, such as notochord, muscle cells and nephric tubules. In recombinates of presumptive ectoderm of the crescent-shaped blastopore stage (stage \(10\frac{1}{2}\)) poor differentiation of mesodermal structures was found; only mesenchyme and blood cells were formed (in 7 out of 20 cases). Moreover, in some recombinates only mesenchymal cells were formed (in 2 out of 20 cases). In some recombinates of the horse-shoe-shaped blastopore stage (stage 11) there was no mesodermal differentiation at all; recombinates consisted only of atypical ectoderm and an undifferentiated endodermal mass (5 out of 20 cases).
Table 1. The reactive capacity of the presumptive ectoderm of different stages comblned with presumptive endoderm of late b1astu1a. Stages according to Nieuwkoop and Faber (1967).
| Recomblnates with presumptive ectoderm of: | |||||
|---|---|---|---|---|---|
| Number of recom binates contain ing: | Early blas tula | Late blas tula | Initial gastrula | Crescent shaped blasto pore | IIorse shoe shaped blas to pore |
| (st.6\/7) | (st.9-) | (st.10) | (s t. 10tJ | (st.11) | |
| 20* | 25* | 20* | 20* | 20* | |
| Fni dpmi e | 20 ( 1002) | 25 (rc0%) | 20 (r00%) | 20 ( 1002) | 20 (1002) |
| Neural tissue | 1 3 ( 652) | 19 ( 76%) | L 2 ( 607") | 7 ( 351!) | 5 ( 2s7!,) |
| Noto chord | 9 ( 4s7") | 10 ( 40%) | 6 ( 30%) | J ( Ls7!,) | 1 ( s%) |
| Mus c1e | 20 (1ooz) | 25 (1oo%) | 1 7 ( 8s%) | 10 ( so7") | 5 ( 257") |
| Nephric tubules | 9 ( 4s%) | 4 ( 167!) | z ( loz) | J ( $%) | 2 ( r}"t) |
| Blood cells | 1 s%) 1 | 8 ( 322) | 12 ( 60%) | L 4 ( 707") | ( 3s%) |
| Mesenchyme | 19 ( es%) | 25 (100%) | 19 ( es%) | 19 ( esz) | 10 ( so7") |
| Meso thelium | 8 ( 4o%) | 19 ( 762) | T ] ( 8s%) | 'J.4 ( 70%) | 5 ( 25"1) |
| Melanophores | 6 ( 30%) | I 7 ( 68%) | J.J ( 6sz) | 12 ( 607") | ( 102) |
| Undifferent iated endoderm | 20 (100%) | 25 (r_002) | 20 (100%) | 20 (100%) | 20 (1002) |
Number of available recombinates
DISCUSSION AND CONCLUSIONS
In the analysis of mesoderm formation in a urodele, axolotl, Nleuwkoop (1969 a) came to the concl-usion that the mesoderm is derived from the ectodermal part of the egg, under the inductl-ve influence of the endoderm. The conclusions drawn from the quantitatlve findings of the experiments on Xenopus Laeuis blastulae correspond with those based on the study of axolotl (Sri Sudarwari and P.D. Nieuwkoop, 1971).
Isolates of the presumptive ecto-neuroderm formed irregu-1ar masses of atyplcal ectoderm in .\lenopus Laeuis, whicl.r corroborates Holtfreterfs findings (1938). As was already found by Ogi (L967) and confirmed by Nieuwkoop (1969), isolarion of the central endodermal mass does not lead to any cellular differentiation in urodeles. This also holds fot Xenopus Laeuis. When the central yolk mass is recornbined with ectodermal cap, the mesoderm 1s newly induced (Sri Sudarwati and P.D. Nieuwkoop, 1971),
Taklng lnto consideration the differedtiation of the main mesodersral structures such as notochord, muscle, nephric tubules and blood cells, the present study reveals that before gastrulatlon the reactive capaclty of the presumptive ectoderm is mainly the production of dorsal mesodermal structures. During the gastrulation process there is a tendency to shift the dlfferentlation of dorsal to ventral mesodermal sEructures. This latter flnding ls In agreement wlth the result of the lnvestlgatlon done by Tseng (1963) on heterogenous inducti.on, using urodelean ectoderm as reactive tissue and euineapig bone marrow as mesoderrnal inductor.
Yamada (1937, l-940) suggests that there is a mesodermal factor whlch controls the differentlatlon of the various inesodermal structures. The present lnvestigation on the reactive capacicy of the presumptlve ecto-neuroderm with increase in age gives supplementary information to the matter. It is probable that the ectoderrnal ce11s before gastrulation possess the reactive capacity to form all the mesodermal differentiations. It may develop into dorsal as well as lateral and ventral structures under the action of a high concerltration of ttrat factor. However, tire presence of large amounts of notochord and muscle suppresses the differentiation of the lateral and ventral mesoderm. Accompanying the increase in age, the ectodermal ce11s rnay undergo certain changes of permeability or biosynthetic activities, so that they can no longer b e acted upon by the ,mesodermal factor at a high concentration. tr^Ihen the condltion to differentiate into notochord and muscle is not fu1fi11ed, the ectodermal ce11s mainly form lateral and ventral mesodermal structures.
The present study also reveals thar the presumptive eetoneuroderm in Xenopus Laeuis has already a competence for mesoderrn formation as early as the early blastula stage (stage
7_, Nieuwkoop and Faber, 1967). The blastula ectoderm (stage 9_) has the highest mesodermization competence. During gastrulation there is a continuous decrease in the capacity of the ectoderm to form mesodermal structures. Both the size and frequency of the structures formed generally decrease with the age of the ectoderm. There was not only a general decrease in reactivity, but also a qualitative change towards epidermal formation. Mesodermization competence starts to fade out in recombinates with ectoderm of the horse-shoe-shaped blastopore stage (stage 11).
SUMMARY
- 1. Mesoderm formation of Xenopus laevis (Daudin) has been studied in order to gain a better insight to the process in amphibians.
- 2. The reactivity of the presumptive ecto-neuroderm in meso-derm formation at various embryonal stages, starting with the early blastula (stage \(6\frac{1}{2}/7\), Nieuwkoop and Faber, 1967) to late gastrula (stage 11) has been studied.
- 3. With the increase in age of the presumptive ecto-neuroderm used in the experiments, there is a shift in the differentiation of the dorsal to ventral mesodermal structures.
- 4. The size and frequency of the occurrence of mesodermal structures generally decrease with the increase in age of the presumptive ecto-neuroderm.
- 5. There was not only a general decrease in reactivity, but also in qualitative change towards epidermal formation.
- 6. The presumptive ecto-neuroderm has already a competence for mesoderm formation as early as the early blastula stage and this competence starts to fade out at the horse-shoe-shaped blastopore stage.
ACKNOWLEDGEMENT
The author is grateful to the Director of the Hubrecht Laboratory, International Embryological Institute, Utrecht-Holland, Prof. Dr. P.D. Nieuwkoop for his guidance, advice and interest in this study. A research grant provided by the Hubrecht Laboratory and The Bandung Institute of Technology is gratefully acknowledged.