Gene commitment
101
Homeotic-like genes in vertebrates: analysis of a gene that is expressed
during early Xenopus development
A. E. Carrasco*, M. MiillerandE.
M. De Robertis, Department of Cell Biology, Biocenter, University of
Basel, Switzerland
Drosophila melanogaster is the only organism in which, after 70 years of genetics, genes that control
animal development have been identified. Using as a probe a region of homology shared by several
Drosophila homeotic genes described by W. McGinnis, W. J. Gehring and colleagues (Nature, 308,428) we
cloned from a Xenopus laevis library two DNA segments that hybridize to the Antennapedia, Ultrabithorax
and fushi tarazu genes of Drosophila. The frog 'homeotic domain' of one gene was sequenced, and it codes
for a peptide region of 60 amino acids that is extremely homologous to the fly genes. The frog conserved
domain represents an 'average' of its three Drosophila counterparts. In other words, if one asks how many
of the amino acids are absent from a particular position in any of the three fly homeotic genes, 58 out of 60
amino acids are conserved. The codon usage provides strong evidence that this sequence must be translated
in the frog. The homology region starts and ends abruptly (not due to introns) and is very arginine rich.
This single-copy frog gene is transcribed into three alternative transcripts detectable by Northerns. One
appears at late gastrulation, peaks during neurulation and then dissappears; a second transcript also starts at
gastrulation but remains elevated even in swimming tadpoles; a third transcript appears at the end of
neurulation. These intriguing genes could turn out to be the first development-controlling genes identified in
vertebrates.
Mitogenic regulation of the commitment of myoblasts to terminal muscle
differentiation
5. Hauschka*, C. Clegg, R. Lim, T. Linkhart, J. Chamberlain, J. Jaynes, C. Bulinski and G. Merrill,
Department of Biochemistry, University of Washington, Seattle, WA 98195, U.S.A.
Terminal differentiation of MM14 mouse myoblasts is regulated by a repression mechanism mediated by
fibroblast growth factor; FGF prevents myogenesis and stimulates proliferation. Other mitogens such as
EGF, PDGF, MSA and insulin do not affect this regulation. Further studies have shown that FGF
repression is direct (i.e., FGF represses differentiation even in non-cycling cells), and that the regulation of
terminal differentiation and DNA synthesis occur exclusively in Gi. Proof that regulation occurs in Gi is
based on the observation that mitotic cells plated into FGF-free medium divide and then differentiate
without further replication. DNA synthesis is thus not required for cells to become competent to
differentiate.
Differentiation occurs within 3 h of FGF removal and isfirstdetected as a commitment response in which
cells exhibit a post-mitotic phenotype which is refractory to subsequent mitogen stimulation. Within 2 h of
this, myocytes have accumulated detectable M-creatine kinase mRNA, and within 3-6 h myocytes stain
positively for acetylcholine receptor, myosin heavy chain (MHC), creatine kinase and o-actin; concurrently,
two myoblast proteins, EGF receptor and thymidine kinase (TK), disappear. While acquisition of the
post-mitotic phenotype generally precedes expression of muscle genes, this may not indicate a causal
relationship, because under specific conditions 0-1-1 % of the cells exhibit MHC and replicate DNA
simultaneously. Thus it appears that following FGF removal activation of the post-mitotic state and
muscle-specific genes occur along independent pathways.
To determine whether differentiation is mediated via the appearance of a diffusible regulator, myocytes
were fused with Gx myoblasts or with various non-myogenic cells and assayed for DNA synthesis. (Myocyte
x Gi myoblast) heterokaryons - even at myocyte:myoblast nuclear ratios of 1:3 - failed to replicate DNA in
response to FGF-rich medium, whereas in (myocyte xGl or even quiescent non-myogenic cell) heterokaryons the myocyte nucleus re-expressed TK and replicated DNA. These studies suggest a model for
the regulation of myogenic differentiation in which a component from an FGF-mediated replication signal
represses differentiation; while in the absence of FGF, differentiation is activated and a diffusible factor is
produced which counteracts the myoblast mitogenic response system - thus conferring the post-mitotic
myocyte phenotype.
102
Gene commitment
Molecular cloning and sequence analysis of repetitive DNA sequences
from Xenopus laevis
W. Knochel, S. Hummel, E. Korge, B. Tappeser and W. Meyerhof, lnstitutfur Molekularbiologie und
Biochemie, Freie Universitdt Berlin, Arnimallee22, D-1000 Berlin 33, Germany
Digestion of Xenopus laevis genomic DNA with Hind III or Eco RI and electrophoretic separation of
restriction fragments shows visible bands of ~0-75 or 0-5 kbp, respectively, after ethidium bromide staining.
Fragments within these two bands were cloned in plasmid pBR 322. Analysis of the cloned fragments by
hybridization studies, restriction maps and by nucleotide sequencing has revealed that (i) the band obtained
by Hind III digestion contains two different major components and (ii) the band obtained by digestion with
Eco RI is mainly composed of three different types of sequences. Three out of 5 sequences (2xHind III,
lxEco RI) are derived from satellite DNAs. One sequence element is dispersed throughout the genome,
while the fifth sequence element is partly organized in tandem arrangement and partly scattered within the
genome. Chromosomal localization of the cloned elements has been investigated by in situ hybridization.
One element is found to be located on a single chromosomal locus, another one seems to be preferentially
located on specific chromosomal loci, while the others are found on most chromosomes. Transcriptional
behaviour of the cloned sequences has been investigated by injection into frog oocytes or by Northern blot
analysis.
Chromosome replication in early Xenopus embryos
R. A. Laskey* (Cambridge)
No abstract for publication
Gene commitment
103
Genetic analysis of cell cycle regulation of HO transcription in yeast
K. A. Nasmyth* andL. L. Breeden, MRC Laboratory of Molecular Biology, Hills Road,
Cambridge CB2 2QH
The diploidization of Saccharomyces cerevisiae is achieved by a very specific pattern of mating type
switching, which involves a double stranded break at the MAT locus by the HO endonuclease. The pattern
of switching may be explained by the pattern of HO transcription, which occurs for a short period late in Gl
in a or or mother cells but is absent throughout the cell cycle of a of a daughter cells and zloc diploids. The
absence of HO in daughters and diploids explains why they never switch. The cell cycle dependence of HO
transcription in haploid mother cells may also explain why switching never occurs during mating and why the
switch is always inherited by both mother and daughter in the next cell division.
In order to identify the cis-acting components of HO regulation, we have carried out an extensive deletion
analysis of nearly 2 kilobase pairs of DNA upstream from the HO gene. We have identified three distinct
regions within this sequence that have important regulatory properties. At about -80 base pairs from the
transcription start site there is a TATA' like sequence that is necessary for transcription and affects the site
of initiation. A second region, more than a kilooase upstrea, must also be present in order to see any HO
transcription at all. In between these two sites there is a region of about 800 base pairs that is involved in
correct cell cycle regulation of HO. Within this region there is a sequence: CACGAAA, which is repeated
eight times. Large deletion mutants lacking these sequences result in constitutive HO transcription early in
the cell cycle. When these sequences are restored by inserting an oligonucleotide bearing this sequence into
deletions that lack it, correct cell cycle regulation of HO transcription is also restored.
Studies are currently underway to identify trans-acting gene products that recognize these CACGAAA
sites and linke the expression of HO to early events in the cell cycle.
Regulation of cell cycle during early Xenopus development
J. Newport* (San Diego)
No abstract for publication
104
Gene commitment
Immunoglobulin class and subclass distributions of antibody responses
to glycoproteins, proteins and carbohydrates in the hamster
/. M. Outschoorn, M. T. GimenezandS. Sanchez Robles, Departments of Preventive Medicine and
Biochemistry, Facultad de Medicina, Universidad Autonoma de Madrid, Madrid 34, Spain
Hamsters were immunized with human erythrocytes of various blood groups, with serum albumins (of
chicken, bovine and human origin) and with a variety of carbohydrates: either commercial polysaccharides
or samples extracted and/or purified from bacteria or fungi. The effect of adjuvants such as Freund's or
polyacrylamide or Bordetella pertussis were studied along with immunization schedules that permitted
comparison of primary, secondary and tertiary responses. Antibody titres were determined on samples of
whole serum and fractions as obtained by gradual or stepwise acid elution from a column of Sepharose
coupled to Staphylococcal Protein A.
high-titred material distributes between the two known subclasses of IgG. In the Syrian hamster IgG2
major subclass eluting first around pH 5 followed by IgGl. Although, to date, only two clearly separable
IgG subclasses have been described in hamsters, our data permits proposal of a third one eluting at or below
pH 4 from Protein A with measurable antibody activity compared with IgGl. Even a fourth peak, also low
in protein content and without specificity for these antigens, is sometimes eluted. Since most responses to
thymus-dependent antigens localize in IgGl, while responses to thymus-independent TI-2 antigens are
found in the broad IgG2 peak, we can also show that this 'IgG3' can share both serological cross-reactivity as
well as biological functions with IgGl on a similar basis to that occurring in Man. Thus manipulation of the
protein and carbohydrate proportions can alter isotype expression as demonstrated by changes in the
relative amounts of the different subclasses synthesized.
Induction and expression of myogenic differentiation using serum-free
medium. Role of the cell cycle
Christian Pinset*, Danny McCormick, Christine Laurent, Gillian S. Butler-Browne and Robert G. Whalen,
Dipartement de Biologie MoUculaire, Institut Pasteur, 25 Rue du Dr. Roux, 75724 Paris, France
In permissive cell culture conditions, expression of the myogenic programme and cell proliferation are
mutually exclusive. We have described (Dev. Biol. 102, 269) conditions which are non-permissive for
myoblast differentiation in the rat cell line L6: in Ham's F12 medium plus fetal calf serum, these myoblasts
will grow to confluence and become arrested in the G0/Gi phase of the cell cycle, but they do not fuse.
Under these conditions, neither synthesis of muscle contractile proteins nor accumulation of the messenger
RNAs for myosin subunits is detectable. Therefore, cessation of DNA synthesis and arrest of cell
proliferation are not sufficient conditions to trigger myogenic diferentiation in these cells. These quiescent
myoblasts are able to re-enter the cell cycle and subsequently fuse and differentiate if the medium is changed
to Dulbecco's MEM (DMEM) containing serum. Thus the absence of differentiation in F12 medium is not
due to an irreversible loss of myogenic potential. Expression of differentiation can also be induced in a
serum-free medium composed of DMEM plus insulin. In this medium, labelling with radioactive thymidine
demonstrates that 90 % of the cells which fuse do so without passing through another S phase. This
conclusion was confirmed by inducing fusion with DMEM plus insulin in the presence of cytosine
arabinoside: accumulation of specific messenger RNAs and synthesis of contractile proteins was initiated
even in the presence of the drug. These experiments represent a direct demonstration that DNA synthesis is
not required for the induction of myogemc differentiation.
Gene commitment
105
Some evidence of cytoplasmic influence on the gene expression of
hybrid fish CtMe obtained from the combination of nucleus and
cytoplasm from two subfamilies of fresh water teleosts,
Ctenopharygodon idellus(Ct) and Megalobrama amblycephala(Me)
Yan Shao-Yi, Wu Naihu, Yan Jingzhi, Xue Guoxiong and Li Guangshan, Institute of Developmental
Biology, Academia Sinica, Beijing, China. Yang Yongquan, Chang Jiang Fisheries Research Institute, Shashi
Branch, Shashi, China
In this paper we report the experiments on the hybrid fish obtained from the combination of nucleus and
cytoplasm from two families of fresh water teleosts by nuclear transplanting method. Nuclei of Ct and
enucleated eggs of Me were used as materials. Both of them belong to the same family, Cyprininae, but Ct
belongs to subfamily, Leucinae and Me belongs to subfamily, Abramidinae.
Starch and polyacrylamide gel electrophoresis as well as immunoelectrophoresis were used for analysing
the Hb, LDH isozymes and serum proteins of Ct, Me and CtMe. The results showed that the Hb
electropherogram of CtMe (9 bands) is different from those of CT (4 bands) and Me (4 bands). The LDH
isozymes electropherograms of gill and kidney of CtMe (5 bands) were different from those of Ct (6 bands)
and Me (5 bands). The serum electropherogram of CtMe was also different from those of Ct and Me in their
band numbers. The serum immunoelectrophoresis showed that the precipitation bands produced in the
combination of CtMe serum and anti-C/Afe serum rabbit serum (15 bands) were different from the
combination of Ct serum and anti-CtMe serum rabbit serum (8 bands) and the combination of Me serum and
anti-CM/e serum rabbit serum (7 bands).
All the results obtained indicate that cytoplasm can influence the expression of genes coding for several
kinds of proteins on the nuclear transplanted hybrid fish.
Paternal gene expression oiXenopus borealis specific protein Nt in
interspecies hybrids between X. laevis and X. borealis
Doris Wedlich, Christine Dreyer* and Peter Hausen, Max-Planck-Institut fiir Entwicklungsbiologie,
Spemannstr. 351V, 7400 Tubingen, West Germany
An abundant acidic germinal vesicle protein of 100000 Da has been described in X. laevis and termed Nx
(1). It is supposed to bind stored histones in the oocyte (2).
Species-specific monoclonal antibodies (mABs) have been raised against the oocyte nuclear protein of X.
borealis, that is equivalent to protein Nx of X. laevis. These mABs have been used to monitor paternal gene
expression of Nx in hybrids between X. laevis and X. borealis.
Protein Nx is accumulated in oocyte nuclei, shed into the cytoplasm of the egg upon germinal vesicle
breakdown, and reaccumulated by the nuclei of the embryo. With development, it appears to be gradually
diluted in all cells of the embryo, its levels falling below the limits of detection after stage 50. In interspecies
hybrids, the paternal antigen is not found in somatic cells, as judged by immunohistological criteria. We
therefore conclude that protein Nx is not expressed from the genes of the embryo and that the maternal
store of Ni is sufficient to endow the nuclei of the embryo up to feeding tadpole stages. This deduction is
corroborated by radiolabelling experiments, and by the observation that paternal genes are, in general, not
suppressed in such interspecies hybrids (3).
The paternal antigen equivalent to Nj is, however, specifically expressed in the germ line. In hybrids and
in X. borealis it is first detected in the nuclei of oogonia and spermatogonia, but in both sexes it is
undetectable during early meiotic prophase. In female germ cells, accumulation of Nx is resumed at the
beginning of diplotene, concomitant with the onset of oocyte growth. The significance of the observed cell
specificity of Ni during germ cell differentiation is discussed in relation to its postulated function as a histone
storage factor.
1) BONNER, W. M. (1975). /. Cell Biol. 64, 431-437.
2) KLEINSCHMIDT, J. & FRANKE, W. W. (1982). Cell 29, 799-809.
3) DE ROBERTIS, E. M. & BLACK, P. (1979). Devi. Biol. 68, 334-339.
106
Gene commitment
Localization and determination in early embryos of Caenorhabditis elegans
W. B. Wood* (Boulder)
No abstract for publication