Many theories of neoplasia suggest that oncogenic transformations result from aberrations in the control mechanisms which normally regulate growth and differentiation during embryonic development. It has recently become clear that many proto-oncogenes are differentially expressed during embryonic development and may thus be important embryonic regulatory molecules. We report here that the products of two transforming oncogenes int-2 and hst / ks (now called kfgf) can, with different potencies, induce mesoderm formation in isolated Xenopus laevis animal pole explants and stimulate DNA synthesis in mammalian fibroblasts. The results suggest that these proteins may function as mesoderm inducers in mammalian embryogenesis and that similar receptor/signalling pathways may be utilized for developmental and oncogenic processes. Finally, we have shown that the Xenopus assay system used in this study provides a powerful screen for protein factors that are active in development.

One of the principal patterning events in early amphibian embryogenesis is the induction of mesoderm tissue from cells of the animal hemisphere by a signal(s) emanating from the cells of the vegetal hemisphere (Nieuwkoop, 1969; Dale et al. 1985; Gurdon et al. 1985). These signals can be mimicked by proteins which are members of the heparin-binding growth factor family (Gospodarowicz et al. 1987) in particular basic (bFGF) and acidic (aFGF) fibroblast growth factor (Slack et al. 1987; Kimelman & Kirschner, 1987; Slack et al. 1988; Kimelman et al. unpublished data). This demonstrates that known growth factors can function as embryonic morphogens.

Two oncogenes, ks/hst (now called kfgf) (Delli-Bovi et al. 1987,1988; Taira et al. 1987) and int-2 (Moore et al. 1986; Dickson & Peters, 1987), which have recently been identified are related by sequence to aFGF and bFGF, being 40– 60% similar at the amino acid level. Previous studies of the expression patterns of int-2 mRNA in embryonal carcinoma (EC) cells and the early mouse embryo have led several authors to suggest that the encoded protein is an inducing factor (R. Smith et al. 1988; Wilkinson et al. 1988, 1989; Nusse, 1988). In the early mouse embryo, expression is found in the primitive streak from 7 · 5 to 9 · 5 days (Wilkinson et al.1988). Since all germ layers of the embryo (ectoderm, mesoderm and endoderm) arise from the primitive streak, it is here rather than in the primitive endoderm that we would expect to find a mammalian mesoderminducing factor. During later development, mRNA expression is found in the endoderm of the first three pharyngeal pouches, the hindbrain adjacent to the developing otocyst and in the mesenchyme of developing teeth (Wilkinson, 1988,1989). These are not centres of mesoderm formation but are sites where the pathways of differentiation are controlled by epithelialmesenchyme interactions and so there may be a role for INT-2 as an extracellular inducing factor in these tissues.

As far as kFGF is concerned there have not been any studies of its expression in normal embryos. However, the protein is known to be synthesized by certain EC cell Unes where its expression is limited to undifferentiated cells (Heath et al. unpublished data). This suggests a possible role for kFGF in cell-cell interactions in normal development.

Here we report that kFGF and INT-2 proteins can, with different potencies, induce mesoderm in Xenopus laevis animal pole cells and stimulate DNA synthesis in mammalian fibroblasts. These data suggest that these proteins may function as mesoderm inducers in mammalian embryogenesis.

In vitro transcription

The complete cDNA clones for bovine basic fgf, human kfgf and murine int-2 were obtained (kfgf and int-2) or subcloned (bfgf) into plasmid vectors containing bacteriophage SP6 and/or T7 RNA polymerase promoters. cRNAs were transcribed in vitro and purified exactly as described by Krieg & Melton (1984). cRNAs were resuspended in diethyl-pyrocarbonate-treated H2O at a concentration of l mg ml-1 and stored at — 70°C until use.

In vitro translations

Proteins were synthesized by in vitro translation of the specific cRNAs in a mRNA-dependent rabbit reticulocyte lysate (Promega Biotech) supplemented with amino acids (ImM each). 1 – 2 μ g of cRNA were heated in a water bath to 70°C for 3 min and rapidly cooled on ice before adding to the reticulocyte lysate mixture to a final volume of 50 μ l. To obtain radioactively labelled proteins, methionine was replaced with 50 μ Ci of [35S]methionine (Amersham; 1000 Ci mmole-1). When canine pancreatic microsomes were included in the translation they were added 5 min after the initiation of translation at between 0 · 5 and 2 · 5 equivalents as defined by the supplier (Promega Biotech). Translations were incubated for 45 min at 30°C and the reactions terminated by placing the tubes on ice. 5 μ l of translation products were digested with 50 μ g ml-1 RNase A (Boehringer Mannheim) for 10 min at 30°C analysed by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) according to Laemmli (1971) using 14% gels. After electrophoresis, the gels were fixed and stained in 0 · 15% Coomassie blue R250 in 45% methanol/10% acetic acid for 15 min then destained overnight in 20% methanol/6% acetic acid before impregnation with Amplify (Amersham) for 30 min and drying. The gels were autoradiographed at —70°C using Kodak X-Omat AR film.

Mesoderm induction assays

All procedures for the fertilization, handling and dissection of Xenopus laevis embryos were as previously described (Godsave et al. 1988). Translations were diluted 1/4 with halfstrength Normal Amphibian Medium (NAM/2) and sonicated with a Soniprep 150 at one-third microtip maximum to disrupt microsomes. Expiants were cultured in 2-fold serial dilutions in NAM/2 of translation products in a total volume of 15 μ l in Terasaki culture plates as previously described (Godsave et al. 1988). Explants were maintained at 25 °C in a humidified chamber for 2 – 3 days before being scored visually for mesoderm induction as described by Godsave et al. (1988) and were then harvested for histology or for RNA extraction.

RNase protection assay

Four to six explants for each treatment were homogenized in 200 μ l of TNS buffer (100mm-Tris-HCl pH7 · 5; 300mm-NaCl; 2% SDS), digested with 100 μg ml-1 Proteinase K (Boehringer Mannheim) for 30 min at 37 °C, extracted twice with phenol, twice with phenol: chloroform: isoamyl alcohol (25:25:1) and precipitated with ethanol at – 20°C. The pellet was washed with cold 70% ethanol, dried and resuspended in DEPC water. RNase protection analysis using the p21 actin cDNA was carried out exactly as described by Gurdon et al. (1985) using 1 explant equivalent of total RNA.

Mitogenicity assays

Nonradioactive translation products were prepared as described, diluted in Dulbecco’s modified Eagles medium (DMEM) containing 0 · 1% foetal calf serum (FCS) and serially diluted in 24-well tissue culture plates (Linbro) containing 5X104 murine embryo fibroblast C3H10T 1/2 cells in DMEM+0 · 1% FCS which had been seeded 24h previously. After 3h, l μ Ci ml-1 of [3H]thymidine (Amersham; specific activity 75 Ci mmol-1) was added to each well and plates were incubated at 37 °C in a humidified incubator containing 5% CO2 for an additional 15 h. At this time, the medium was removed, the monolayers washed with phosphate-buffered saline (PBS; pH7 · 0) and processed for autoradiography as previously described (Heath, 1987). After 3 days, the plates were developed, stained with 1 % Giesma in PBS and visually scored for positive nuclei.

Heparin-binding assay

Translations were performed as described and 50 μ l samples were diluted to 200 μ l in loading buffer to give a final concentration of 10 mm-Hepes pH7 · 5, 400 mm-NaCl and applied to a heparin sepharose (Pharmacia) column (bed volume 0-3 ml) equilibrated in the same buffer. The column was washed with 5ml of wash buffer (10 mm-Hepes pH7 · 5, 400 mm-NaCl, 0 · 1% haemoglobin) and eluted in 2 × 0 · 25ml volumes of elution buffer (10 mm-Hepes pH7 · 5, 2m-NaCl, 01% haemoglobin). Bound and unbound fractions were desalted using Centricon 10 concentrators (Arnicon) and resuspended to equivalent volumes in 10 mm-Hepes pH7 × 5, 10 mm-NaCl, 0 × 1% haemoglobin. Volumes equivalent to unfractionated lysates were used to stimulate DNA synthesis in mouse fibroblasts described above except that [3H] thymidine incorporation in TCA-insoluble material was solubilized in 0 × 3 M-NaOH, 1 % SDS and counted by liquid scintillation.

We initially attempted to investigate the mesoderminducing activities of bFGF, kFGF and INT-2 by injection of the specific cRNAs into early Xenopus embryos, excising the animal cap at stage 8 and assaying for mesoderm induction as described (Godsave, 1988). We found that with bfgf or kfgf the efficiency of mesoderm induction was poor, never exceeding 25 –30% over control levels regardless of the concentration of injected RNA, the stage at which it was injected or the location of injection (data not shown). We did not consider that such results were suitable for quantitative work.

This led us to try a different method, which was to synthesize the specific protein in a rabbit reticulocyte lysate system and measure its mesoderm-inducing activity by treatment of ectoderm explants with serial dilutions of the lysate. Mesoderm differentiation was assayed by visual inspection of elongation and vesicle formation by the explants (Godsave et al. 1988) followed by either histological examination or measurement of the cardiac-specific form of actin mRNA (a marker gene for mesoderm) by RNAase protection analysis (Gurdon et al. 1985; Mohun et al. 1984). Canine pancreatic microsomes were included in translations to allow for possible post-translational modifications of the proteins which could be important for biological activity.

In each case, the cRNA was efficiently translated to yield proteins of the size predicted from inspection of the DNA sequence (INT-2) or knowledge of the secreted protein (bFGF, kFGF) (Fig. 1). The core proteins of the translation products have apparent molecular weights of 19 500 for bFGF, 26 000 for kFGF and 32 000 for INT-2. In our conditions, we could observe no evidence of post-translational processing of bFGF nor a detectable lower molecular weight product corresponding to removal of a signal sequence for kFGF or INT-2.

Fig. 1.

SDS-polyacrylamide electrophoresis (SDS-PAGE) of 35S-Iabelled translation products of cRNA encoded by cDNAs to bfgf, kfgf and int-2. Con indicates translation in the absence of added cRNA. The arrows indicate probable TV-linked glycosylation products as these bands do not appear in translations in the absence of canine pancreatic microsomes (data not shown).

Fig. 1.

SDS-polyacrylamide electrophoresis (SDS-PAGE) of 35S-Iabelled translation products of cRNA encoded by cDNAs to bfgf, kfgf and int-2. Con indicates translation in the absence of added cRNA. The arrows indicate probable TV-linked glycosylation products as these bands do not appear in translations in the absence of canine pancreatic microsomes (data not shown).

The MIF assays clearly showed that bfgf and kfgf and int-2 cDNAs all encode proteins with mesoderm-inducing activity whether measured by explant morphology (Table 1) or expression of a muscle-specific gene (Fig. 2). Nevertheless, significant qualitative and quantitative differences between the effects of the three proteins were apparent. Based upon the calculated concentration of protein synthesized in the reticulocyte system (see Table 1), both bFGF and kFGF are potent mesoderm-inducing factors possessing specific activities of l · 3 × 107U mg-t and 3 concentration of protein synthesized in the reticulocyte 2 × 106U mg-1 respectively, where 1 unit of inducing activity is the amount of 1 ml of medium that will provoke vesicle formation in 50 % of the explants (Godsave et al. 1988). INT-2 was found to be significantly less potent in this assay; mesoderm induction was only observed at relatively high concentrations of INT-2 protein (specific activity = 2 · 0 × 105 U mg-1). It should be noted, however, that the specific activity measurements were calculated with the assumption that the core and processed forms of kFGF and INT-2 have the same specific activity.

Table 1.
graphic
graphic
Fig. 2.

RNase protection analysis of a cardiac musclespecific actin mRNA in explants treated with in vitro translation products. Explants were exposed to dilutions of translation cocktail equivalent to 16 MIF units ml-1 of bFGF, 16 MIF units ml-1 of kFGF and 2 MIF units ml-1 of INT-2 and an equivalent dilution of control lysate for 3 days after which time RNA was extracted for analysis. The RNA probe protects a cardiac-specific actin RNA fragment of 285 bases and a cytoskeletal actin fragment of approximately 235 bases. The cytoskeletal bands serve as internal controls for the amount of RNA per lane. ST 28 refers to total RNA extracted from intact stage-28 embryos and serves as a control for cardiac actin mRNA expression.

Fig. 2.

RNase protection analysis of a cardiac musclespecific actin mRNA in explants treated with in vitro translation products. Explants were exposed to dilutions of translation cocktail equivalent to 16 MIF units ml-1 of bFGF, 16 MIF units ml-1 of kFGF and 2 MIF units ml-1 of INT-2 and an equivalent dilution of control lysate for 3 days after which time RNA was extracted for analysis. The RNA probe protects a cardiac-specific actin RNA fragment of 285 bases and a cytoskeletal actin fragment of approximately 235 bases. The cytoskeletal bands serve as internal controls for the amount of RNA per lane. ST 28 refers to total RNA extracted from intact stage-28 embryos and serves as a control for cardiac actin mRNA expression.

An unexpected observation from these experiments was that bovine bFGF synthesized in vitro was more potent as a mesoderm-inducing agent than bFGF purified from bovine brain. Not only was the titre 10 times higher but histological examination revealed that a significant proportion of the induced explants contained notochord, a dorsal mesodermal structure which has not been observed in explants exposed to similar concentrations of brain-derived bFGF (Slack et al. 1987,1988). Explants treated with kFGF also contained dorsal mesodermal tissues, including notochord, in a similar proportion of explants as bFGF. In no case was dorsal mesoderm differentiation observed in explants exposed to INT-2 protein emphasizing its inferior potency in this assay. We considered the possibility that the enhanced potency of bFGF synthesized in vitro was due to synergistic interaction with other factors present in the translation reaction mixture. However, purified brain-derived bFGF was equipotent as a mesoderminducing factor whether exposed to explants in simple salt solution (NAM/2) or in the presence of reticulocyte lysate, and no induction of notochord was observed in either situation. It should be noted, however, that some batches of reticulocyte lysates can contain low levels of endogenous mesoderm-inducing activity and components which may act synergistically with added bFGF. Nevertheless, these results indicate that bFGF synthesized in vitro is both quantitatively and qualitatively more potent than its ‘natural’ brain-derived equiv-aient. This may result from some hitherto unsuspected inactivation phenomenon arising from the conditions of bFGF purification from brain or the fact that bFGF synthesized in vitro is somewhat larger than purified brain bFGF (data not shown), probably resulting from the continued presence of 9 TV-terminal amino acids in the in vitro translated product which are absent in the most abundant bFGF form purified from natural sources (Abraham et al. 1986).

The best known effect of bFGF is its ability to act as a mitogen for cells of mesenchymal origin (Gospodarowicz et al. 1987) while kFGF expressed under the control of a heterologous promoter in COS cells has also been shown to have mitogenic properties equivalent to bFGF (Delli-Bovi et al. 1988). However, the mitogenic properties of the int-2 gene product have not previously been investigated. We accordingly tested bFGF, kFGF and INT-2 synthesized in vitro for the ability to induce DNA synthesis in C3H10T1/2 mouse fibroblast cells (Table 2). Both bFGF and kFGF were found to be potent mitogens in this system with activities detectable at concentrations of 5 – 50 pg ml-’ (data not shown). INT-2 synthesized in vitro was also able to induce DNA synthesis albeit with an approximately 10fold lower relative activity.

Table 2.
graphic
graphic

A characteristic feature of FGF-like growth factors is their affinity for heparin so we tested the ability of all three proteins to stimulate DNA synthesis after fractionation of the translation mixtures on heparin sepharose columns (Table 3). The results show that the bFGF and kFGF proteins were able to bind heparin but INT-2 was not, suggesting an important functional distinction between INT-2 and other members of the FGF family.

Table 3.
graphic
graphic

One of the most striking observations of these results is that the mitogenicity of these proteins parallels their mesoderm-inducing potencies suggesting that similar receptor/signalling pathways are involved in these two distinct processes. The low specific activity of INT-2 in both assays may indicate that it is intrinsically less active than other members of the FGF family, or that some form of post-translational modification, which is not performed in our in vitro system, is required for full biological activity. A final possibility is that INT-2, being the least similar in amino acid sequence to the prototype FGF molecule (Dickson & Peters, 1987), can only cross-react with the FGF receptor at relatively high concentrations, and perhaps a second high-affinity receptor system for INT-2 exists which is either not expressed in the cell types examined here, or does not mediate either mesoderm-inducing or mitogenic signalling effects.

This study is the first to show any biological activity for int-2, a gene which has aroused great interest by virtue of its involvement in viral oncogenesis and its spatial pattern of expression in early mammalian embryogenesis. There is no satisfactory mammalian version of the serial dilution assay for MIF activity but we feel that its activity as a MIF in the Xenopus assay strengthens the case for it being a mesoderm-inducing factor in early mammalian development as well.

Furthermore, we have found that the in vitro transcription/translation system system coupled with the serial dilution assay for MIFs provides a simple, rapid, visual and sensitive method for assaying active mesoderm inducers. We feel that this assay is not limited to the investigation of amphibian development since it provides a test for factors which now include not only all the members of the FGF family examined to date but some members of the TGF β family as well (Smith, 1987; J. C. Smith et al. 1988; Rosa et al. 1988).

We are grateful to Give Dickson and Gordon Peters (ICRF) for the int-2 clone, Dr T. Mohun for the actin cDNA, Dr C. Basifico (NYU Medical School) for the kfgf clone and to Dr J. Abraham (CalBio) for permission to use the bfgf clone. We thank Sue Godsave for critical reading of the manuscript. LLG is a recipient of a Centennial Fellowship from the MRC of Canada. JKH is supported by the Cancer Research Campaign and GDP, MD and JMWS by the Imperial Cancer Research Fund.

Abraham
,
J. A.
,
Mergla
,
A.
,
Whang
,
J. L.
,
Tumulo
,
A.
,
Friedman
,
J.
,
Hjerrild
,
K. A.
,
Gospodarowicz
,
D.
&
Fiddes
,
J. C.
(
1986
).
Nucleotide sequence of a bovine clone encoding the angiogenic protein, basic fibroblast growth factor
.
Science
233
,
545
548
.
Dale
,
L.
,
Smith
,
J. C.
&
Slack
,
J. M. W.
(
1985
).
Mesoderm induction in Xenopus laevis: a quantitative study using a cell lineage label and tissue-specific antibodies
.
J. Embryol. exp. Morph
.
89
,
289
313
.
Delli-Bovi
,
P.
,
Curatola
,
A. M.
,
Kern
,
F. G.
,
Greco
,
A.
,
Ittmann
,
M.
&
Basilico
,
C.
(
1987
).
An oncogene isolated by transfection of Kaposi’s sarcoma DNA encodes a growth factor that is a member of the FGF family
.
Cell
50
,
729
737
.
Delli-Bovi
,
P.
,
Curalota
,
A. M.
,
Newman
,
K. M.
,
Sato
,
Y.
,
Moscatelli
,
D.
,
Hewick
,
R. M.
,
Rifkin
,
D.
&
Basilico
,
C.
(
1988
).
Processing, secretion and biological properties of a novel growth factor of the fibroblast growth factor family with oncogene potential
.
Mol. Cell. Biol
.
8
,
2933
2941
.
Dickson
,
C.
&
Peters
,
G.
(
1987
).
Potential oncogene product related to growth factors
.
Nature, Lond
.
326
,
833
.
Godsave
,
S. F.
,
Isaacs
,
H.
&
Slack
,
J. M. W.
(
1988
).
Mesoderminducing factors: a small class of molecules
.
Development
102
,
555
566
.
Gospodarowicz
,
D.
,
Ferrara
,
N.
,
Schweigerer
,
L.
&
Neufeld
,
G.
(
1987
).
Structural characterization and biological functions of fibroblast growth factor
.
Endocrine Rev
.
8
,
95
114
.
Gurdon
,
J. B.
,
Fairman
,
S.
,
Mohun
,
T. J.
&
Brennan
,
S.
(
1985
).
Activation of muscle-specific actin genes in Xenopus development by an induction between animal and vegetal cells of a blastula
.
Cell
41
,
913
922
.
Heath
,
J. K.
(
1987
).
Experimental analysis of teratocarcinoma cell multiplication and purification of embryonal carcinoma cell derived growth factor
.
In Teratocarcinomas and Embryonal Stem Cells: A Practical Approach
, (ed.
E. J.
Robertson
).
IRL Press
.
Kimelman
,
D.
&
Kirschner
,
M.
(
1987
).
Synergistic induction of mesoderm by FGF and TGF-/J and the identification of an mRNA coding for FGF in the early Xenopus embryo
.
Cel!
51
,
869
877
.
Kimelman
,
D.
,
Abraham
,
J. A.
,
Haaparanta
,
T.
,
Palisi
,
T. M.
&
Kirschner
,
M. W.
(
1988
).
The presence of fibroblast growth factor in the frog egg: its role as a natural mesoderm inducer
.
Science
242
,
1053
1056
.
Krieg
,
P. A.
&
Melton
,
D. A.
(
1984
).
Functional messenger RNAs are produced by SP6 in vitro transcription of cloned cDNAs
.
Nucleic Acids Res
.
12
,
7057
7070
.
Laemmli
,
U.
(
1971
).
Cleavage of structural proteins during assembly of the head of bacteriophage T4
.
Nature, Lond.
227
,
680
685
.
Mohun
,
T. J.
,
Brennan
,
S.
,
Dathan
,
N.
,
Fairman
,
S.
&
Gurdon
,
J. B.
(
1984
).
Cell-type specific activation of actin genes in the early amphibian embryo
.
Nature, Lond
.
311
,
716
721
.
Moore
,
R.
,
Casey
,
G.
,
Brookes
,
S.
,
Dixon
,
M.
,
Peters
,
G.
&
Dickson
,
C.
(
1986
).
Sequence, topography and protein coding potential of mouse int-2: a putative oncogene activated by mouse mammary tumour virus
.
EMBO J
.
5
,
919
924
.
Nieuwkoop
,
P. D.
(
1969
).
The formation of mesoderm in the Urodealean amphibians. 1. Induction by the endoderm
.
Wilhelm Roux Arch. Entwmech. Org
.
162
,
341
373
.
Nusse
,
R.
(
1988
).
The int-2 genes in mammary tumorigenesis and in normal development
.
Trends in Genet
.
4
,
291
295
.
Rosa
,
F.
,
Roberts
,
A. B.
,
Danielpour
,
D.
,
Dart
,
L. L.
,
Sporn
,
M. B.
&
Dawid
,
I. B.
(
1988
).
Mesoderm induction in amphibians: the role of TGF-ft-like factors
.
Science
239
,
783
785
.
Slack
,
J. M. W.
,
Darlington
,
B. G.
,
Heath
,
J. K.
&
Godsave
,
S. F.
(
1987
).
Mesoderm induction in early Xenopus embryos by heparin-binding growth factors
.
Nature, Lond
.
326
,
197
200
.
Slack
,
J. M. W.
,
Isaacs
,
H. V.
&
Darlington
,
B. G.
(
1988
).
Inductive effects of fibroblast growth factor and lithium ion on Xenopus blastula ectoderm
.
Development
103
,
581
590
.
Slack
,
J. M. W.
&
Isaacs
,
H. V.
(
1989
).
Presence of basic fibroblast growth factor in the early Xenopus embryo
.
Development
105
,
147
154
.
Smith
,
J. C.
(
1987
).
A mesoderm inducing factor is produced by a Xenopus cell line
.
Development
99
,
3
14
.
Smith
,,
J. C.
,
Yaqoob
,
M.
&
Symes
,
K.
(
1988
).
Purification, partial characterization and biological effects of the XTC-mesoderm-inducing factor
.
Development
103
,
591
600
.
Smith
,
R.
,
Peters
,
G.
&
Dickson
,
C.
(
1988
).
Multiple RNAs expressed from the int-2 gene in mouse embryonal carcinoma cell lines encode a protein with homology to fibroblast growth factor
.
EMBO J
.
7
,
1013
1022
.
Taira
,
M.
,
Yoshida
,
T.
,
Miyagawa
,
K.
,
Sakamoto
,
H.
,
Terada
,
M.
&
Sugimura
,
T.
(
1987
).
cDNA sequence of human transforming gene hst and identification of the coding sequence required for transforming activity
.
Proc. natn. Acad. Sci. U.S.A
.
84
,
2980
2984
.
Wilkinson
,
D. G.
,
Peters
,
G.
,
Dickson
,
C.
&
Macmahon
,
A. P.
(
1988
).
Expression of the FGF-related proto-oncogene int-2 during gastrulation and neurulation in the mouse
.
EMBO J
.
7
,
691
695
.
Wilkinson
,
D. G.
,
Bhatt
,
S.
&
Mcmahon
,
A. P.
(
1989
).
Expression pattern of the FGF-related proto-oncogene int-2 suggests multiple roles in foetal development
.
Development
105
,
131
136
.