Plenary lectures
Extracellular matrix regulation of epithelial behaviour
in morphogenesis
1
Merton Bernfield*, Shib Banerjee, Markku Jalkanen, Joy Koda, Hung Nguyen and Alan Rapraeger,
Department of Pediatrics, Stanford University Medical Center, Stanford, CA 94305, U.S.A.
The substratum for many cell types is the extracellular matrix (ECM). However, the ECM is not only a
physical supporting framework for the cells, but may be linked to the cells by molecules which are intimately
associated with the plasma membrane. Changes in cell-ECM interactions modify cell shape, motility, and
proliferation and, because the ECM distributes over many cells, can integrate these cellular events into
outcomes at the tissue level. Our work on the mechanisms by which mesenchymal tissues direct the
generation of epithelial organ form is based on the idea that the interacting tissues reciprocally modify the
other's cell-ECM associations. Epithelia of organs such as the lung, submandibular and mammary glands
begin as buds which then, under the influence of their associated mesenchymal tissue, fold to form lobules
and ultimately ducts and acini. Between these tissues is a basement membrane, an ECM consisting of an
epithelially-derived basal lamina and a mesenchymally-derived reticular lamina. The basal lamina, a thin,
organized complex of principally type IV collagen, laminin, and proteoglycan, acts as the epithelial cell
substratum. The reticular lamina, an adjacent, thick, fibrillar layer containing primarily types I, III, and V
collagen and fibronectin, provides support for the basal lamina. These materials modify epithelial cell
behavior in vitro: isolated basal laminae (or purified constituents) promote cell attachment and
proliferation, and, type I collagen fibrils organize cells into structures resembling ducts and acini. We have
used mouse embryo submandibular glands and NMuMG mammary epithelial cells to examine the role of
the ECM in morphogenetic tissue interactions.
Work on the submandibular gland showed that the basal lamina maintains epithelial morphology, but that
the mesenchyme is required for changes in this morphology and led to the proposal that the mesenchyme
dictates changes in epithelial morphology by remodeling the basement membrane. Basement membrane
materials are distributed non-uniformly on the developing epithelium; a well-defined basal lamina surrounds
the pre-lobular bud but reticular lamina components are only on its stalk. These distributions change with
the formation of clefts and lobules; the basal lamina becomes incomplete on the lobules, the sites of rapid
proliferation and collagen fibrils accumulate within the clefts, the sites of morphologic stability. These
changes correlate with the turnover rates of basal lamina glycosaminoglycans (GAG). GAG (chondroitin
sulfate and hyaluronate) is lost from the lobules because its rate of degradation is greater than its rate of
replacement, while other GAG (notably heparan sulfate) accumulates within the clefts. The incomplete
laminae are transient and, with further development, a complete basement membrane develops around
each acinus and duct. These changes in ECM distribution and composition are due, in part, to the
mesenchyme. The mesenchyme degrades basal lamina GAG and contains a soluble, non-lysosomal, neutral
hyaluronidase which is developmentally regulated during submandibular morphogenesis. The mesenchyme
deposits the collagen-rich reticular lamina which causes the accumulation of a basement membrane heparan
sulfate proteoglycan. Thus, the mesenchyme remodels the basement membrane, producing microheterogeneities in its composition and organization which correlate with changes in epithelial cell behavior.
Such remodelling may account for similar microheterogeneities in other emoryonic organs and may be a
general mechanism for regulating cell behavior in tissue interactions.
Epithelia cultured within a network of type I collagen fibrils respond by forming branches, ducts and
acini, suggesting that they have a mechanism for recognizing these fibrils. The recognition and response
appear to involve a cell surface heparan sulfate-rich proteoglycan. Mouse mammary epithelial cells contain
a lipophilic proteoglycan which is distinct from their basement membrane proteoglycan in solubility, size,
buoyant density and immunoreactivity. This proteoglycan is an integral membrane protein whose GAG
chains are exposed at the cell surface and is mobile within the plane of the membrane. The proteoglycan is
anchored to the cell by linkage either intracellularly or to other integral membrane protein(s), but is rapidly
shed when cell-substratum associations are disrupted. Interactions of the cell surface proteoglycan, studied
using the soluble, non-1
V, but not types II or
of high affinity sites (Kd ca. 10~y M) and is abolished by heparinproteoglycan appears to be a cell surface receptor for interstitial collagen; intact cells mimic its collagen
binding properties and its mobility in the plane of the membrane is blocked by type I collagen fibrils. The
branching induced by culturing mammary epithelial cells within type I collagenfibrilsis rapidly abolished by
heparin. Thus, the proteoglycan links the epithelial cells to the collagen fibrils, potentially translating ECM
organization into cell orientation and cell shape.
(Supported by grants HD06763 and CA28735 from the National Institutes of Health and BC-409 from the
American Cancer Society.)
2
Plenary lectures
Localized expression of homeotic genes and the control of
embryogenesis in Drosophila
Walter J. Gehring*, Biozentrum, University of Basel, CH-4056 Basel, Switzerland
In Drosophila cell determination takes place in the early embryo when the cleavage nuclei migrate to the
surface of the egg and a monolayer of cells, the blastoderm, is formed. The cleavage nuclei are still
totipotent, but the blastoderm cells are determined to give rise to specific segments. At least 3 classes of
genes are involved in the control of early cell determination: Maternal effect genes which are active during
oogenesis and involved in the spatial organization of the egg, zygotic genes which affect the number and
polarity of the segments (segmentation genes) and homeotic genes which specify segment identity.
We have cloned the homeotic Antennapedia (Antp) locus which is necessary for the proper differentiation
of the mesothoracic segment by 'walking along the chromosome' (Garber et al. 1983; Levine et al. 1983).
The analysis of Antp cDNA clones revealed some crosshomology between various homeotic genes
(McGinnis et al. 1984a). The homology is confined to a short DNA segment of 180 basepairs, the homeo
box, which codes for a highly conserved protein sequence (McGinnis et al. 19846; Carrasco et al. 1984). On
the basis of this crosshomology we have isolated several other homeotic genes including P99, which has
tentatively been identified as the Deformed (Dfd) locus (McGinnis et al. 1984a) and fushi tarazu (ftz)
(Kuroiwa et al. 1984), a segmentation mutant which lacks alternate segments.
Using an improved method of in situ hybridization which we developed for the detection of rare
transcripts in frozen sections (Hafen et al. 1983), we have studied the localization
of ftz and Dfd transcripts
at early embryonic stages in the wildtype. At the blastoderm stage ftz+ transcripts are localized in a well
defined pattern of seven segmental bands along the anterior-posterior axis of the embryo (Hafen et al.
1984). This localization can first be detected prior to the formation of cell membranes during the syncytial
blastoderm stages. The Dfd+ transcripts are confined to a single band across the blastoderm and gastrula, in
the posterior head region. The spatial distribution of these transcripts reflects the state of determination of
the blastoderm nuclei. We propose that the cleavage nuclei become determined by an interaction with
determinants in the egg cortex which specify the position within the egg, and that genes like ftz and Dfd
serve as 'sensors' for these determinants.
A. E., MCGINNIS, W., GEHRING, W. J. & DEROBERTIS, E. M. (1984). Cloning of a Xenopus
laevis gene expressed during early embryogenesis that codes for a peptide region homologous to
Drosophila homeotic genes: implications for vertebrate development. Cell (in press).
GARBER, R. L., KUROIWA, A. & GEHRING, W. J. (1983). Genomic and cDNA clones of the homeotic locus
Antennapedia in Drosophila. The EMBO J. 2, 2027-2036.
HAFEN, E., LEVINE, M., GARBER, R. L. & GEHRING, W. J. (1983). An improved in situ hybridization method
for the detection of cellular RNAs in Drosophila tissue sections and its application for localizing
transcripts of the homeotic Antennapedia gene complex. The EMBO J. 2, 617-623.
HAFEN, E., KUROIWA, A. & GEHRING, W. J. (1984). Spatial distribution of transcripts from the segmentation
gene fushi-tarazu during Drosophila embryonic development. Cell (in press).
KUROIWA, A., HAFEN, E. & GEHRING, W. J. (1984). Cloning and transcriptional analysis of the
segmentation gene fushi tarazu of Drosophila. Cell (in press).
LEVINE, M., HAFEN, E., GARBER, R. L. & GEHRING, W. J. (1983). Spatial distribution of Antennapedia
transcripts during Drosophila development. The EMBO J. 2, 2037-2046.
MCGINNIS, W., LEVINE, M. S., HAFEN, E., KUROIWA, A. & GEHRING, W. J.'(1984a). A conserved DNA
sequence in homeotic genes of the Drosophila Antennapedia and bithorax complexes. Nature 308,
428-433.
MCGINNIS, W., GARBER, R. L., WIRZ, J., KUROIWA, A. & GEHRING, W. J. (19846). A homologous
protein-coding sequence in Drosophila homeotic genes and its conservation in other metazoans. Cell (in
press).
CARRASCO,
Plenary lectures
Glycoproteins and cell adhesion in Dictyostelium
3
G. Gerisch*, J. Stadler, G. Bertholdt, K. Toda and S. Bozzaro, Max-Planck-Institut fur Biochemie,
D 8033 Martinsried, Germany
Aggregating cells of D. discoideum are distinguished from growth phase cells by specific membrane
glyoproteins, one of which has been identified as a target antigen of aggregation blocking antibody
fragments (Fab). This glycoprotein with an apparent molecular weight of 80 kd has previously been
designated as contact site A. It is an integral membrane protein modified by highly sulfated carbohydrate
residues, and also by acylation with palmitic acid. Expression of the glycoprotein is controlled by cyclic
AMP which also regulates EDTA-stable contact formation, a characteristic of aggregation competent cells.
Mutants and monoclonal antibodies are used to investigate the developmental regulation of the glycoprotein
and its function.
Cells of a related species, Polysphondylium pallidum, sort out during aggregation from those of D.
discoideum. P. pallidum cells have immunologically distinct glycoproteins with L-fucose at non-reducing
ends of their oligosaccharide chains. Monoclonal antibody directed against these oligosaccharide end groups
completely blocks cell adhesion in P. pallidum.
4
Plenary lectures
Regionalisation in the molluscan embryo
J. A. M. van den Biggelaar*, W. J. A. Arnolds, C. A. M. van Dongen, A. W. C. Dorresteijn and W. M.
Kiihtreiber, Zoological Laboratory, University of Utrecht, The Netherlands
Embryonic and postembryonic differentiation are different. The former leads to cell diversification and
morphogenesis, the latter to proliferation of a single cell type and morphostasis. After its formation a
somatic cell can differentiate directly. The zygote, however, cannot differentiate into whatever kind of cell
without a number of preceding cleavages. In order to produce terminally differentiated (somatic) cells, the
ooplasm has to be sorted out and partitioned between various cell lineages. The specific pattern of
ooplasmic segregation is compartimentilized by a more or less rigid cleavage pattern and it does not only
determine the specification of cell types, but also the organisation of the body plan of the species.
In some species a prepolarity, in the form of a heterogeneous distribution of the ooplasm and differences
in the properties of the egg membrane, can be traced back to the position of the oocyte in the ovary
(Dohmen, 1983; van den Biggelaar and Guerrier, 1983). In species with equally dividing eggs the definitive
polarity may arise independently of the prepolarity. Here the first sign of egg polarity is the eccentric
position of the maturation spindles. Due to the original prepolarity the spindles are attracted to the egg
surface. Then, this polarity is enhanced by the deep aster. The polarizing effect of both maturation asters
can be derived from pressure experiments on oocytes of Umax. If second maturation is forced to become
equal, then in both halves the asters develop the same pattern of ooplasmic segregation, but with an
opposite direction of the polarity and independent of the original egg axis (Guerrier, 1968). The original
polarity, however, may be sufficient: in the absence of maturation spindles vegetal halves of equatorially
subdivided oocytes of Patella develop into normally cleaving embryos the axis of which parallels the original
animal-vegetal polarity (van den Biggelaar, unpublished).
Ooplasmic segregation in cooperation with a well defined cleavage pattern is an essential mechanism for
the restriction of the developmental capacities of the blastomeres but it is not sufficient. It must be
accompanied or followed by cellular interactions. From deletion experiments we know that the micromeres
of the first quartet develop all head structures (van Dam and Verdonk, 1932). Yet the micromeres are not
determined to form a specific part of the head. After deletion of one micromere one may obtain normal
embryos if the gap is filled symmetrically by the three remaining cells (Arnolds, van den Biggelaar and
Verdonk, 1983). This result stresses the importance of the position of a cell for the determination of its final
contribution to the embryo.
Another example of the significance of cell position, and also of cellular interaction, is the determination
of the stem cell of the mesoderm in the Patella embryo. Each of the macromeres at the vegetal pole of the
originally symmetrical 32-cell embryo has the capacity to form mesoderm; only the one that attains a central
position in the embryo and makes contact with the micromeres of the opposite animal pole, however, is
induced to develop mesoderm (van den Biggelaar and Guerrier, 1979; Arnolds et al. 1983). The exchange
mechanism of the signal from the inducer cells to the induced macromere is not known. Gap junctions are
not involved in the transduction of this signal, as far as can be judged from the absence of transport of low
molecular weight substances between these cells (Dorresteijn et al. 1983). We cannot exclude exchange of
substances, as we have obtained embryos in which horse radish peroxidase has been transported between
the central macromere and the involved micromeres (Kiihtreiber, unpublished observations).
An alternative possibility for the induction of the mesoderm is the exchange of a 'message' by surface
molecules. Prior to the inductive period the inducer cells produce an extracellular matrix with which only
the presumptive mesodermal stem cell makes contact (Dorresteijn, unpublished). This extracellular matrix
binds three types of lectins, and presumably anti-fibronectin (Kiihtreiber, unpublished). Introduction of
lectins or anti-fibronectins into the cleavage cavity will be necessary to investigate the role of the
extracellular matrix in the cellular interactions for the induction of the mesoderm.
ARNOLDS,
W. J. A.,
VAN DEN BIGGELAAR,
J. A. M. & VERDONK, N. H. (1983). Roux'sArch. Dev. Biol. 192,
75-85.
VAN DEN BIGGELAAR, J. A. M. & GUERRIER, P. (1979). Dev. Biol. 68, 462-471.
VAN DEN BIGGELAAR, J. A. M. & GUERRIER, P. (1983). The mollusca (ed. K. M. Wilbur),
vol. 3 Development
by N. H. Verdonk, J. A. M. van den Biggelaar & A. Tompa, pp. 179-213. Academic Press, N.Y.
VAN DAM, W. I. & VERDONK, N. H. (1982). Roux's Arch. Dev. Biol. 191112-118.
DOHMEN, M. R. (1983). The mollusca (ed. K. M. Wilbur), vol. 3 Development^ N. H. Verdonk, J. A. M.
van den Biggelaar & A. Tompa, pp. 1-48. Academic Press, N. Y.
DORRESTEIJN, A. W. C , WAGEMAKER, H. A., DE LAAT, S. W. & VAN DEN BIGGELAAR, J. A. M. (1983).
Roux's Arch. Dev. Biol. 192, 262-269.
P. (1968). Ann. d'Embryol. Morphogen. 1, 119-139.
GURRIER,
Plenary lectures
Rapid genomic change in maize. Is this a new form of developmental
plasticity?
5
Virginia Walbot*, Department of Biological Sciences, Stanford University, Stanford, C.A. 94301, U.S.A.
There are fundamental differences between the lifestyles of higher plants and animals. In higher animals
an essentially complete organism is formed during embryonic development including the setting aside of a
germline which will function late in development to produce gametes. Animals respond to a changing
environment during their adult life by behavior and physiological means, only rarely developing new
structures to cope with environmental stress. Higher plants, on the other hand, are sessile and can rarely use
behavior to escape stress; rather plants depend on physiological and ultimately developmental adaptations.
During plant embryo development the key event is the establishment of the meristems which act as stem cell
populations throughout adult life, continuously producing new organs, including the germ cells under
appropriate conditions. Thus the embryonic plant contains only the first examples of the vast number of
organs that will be produced throughout the plant's life span. Plants are able to produce organs such as
leaves that are specifically adapted to the conditions existing at the time of organ initiation and growth - the
same plant can produce leaves adapted to bright sun or dense shade.
This developmental plasticity of plants is matched by flexibility in the organization of the genome. Higher
plants are much more tolerant of chromosomal changes such as aneuploidy and translocations than higher
animals. Plants with unusual karyotypes are common in nature and can be reproductively successful. In
addition, we have discovered that rapid genomic change is a feature of the maize genome and, by
extrapolation, may be a feature of all angiosperm genomes.
In 1980 we reported that the modern maize genome shows unexpectedly high divergence from primitive
maize and maize relatives. These differences suggested to us that the maize genome could change rapidly on
an evolutionary time scale. Now we have evidence that the genome can change in a single generation. We
believe that these rapid genomic changes are part of the developmental strategy of the plant; that is, a
change in the environment triggers genomic change which in some cases results in cell lineages capable of
forming organs better adapted to local conditions.
Comparing inbred lines of maize we find that each line has afixedcopy number for a variety of repetitive
sequences we have tested; in other words, there are stable polymorphisms for copy number within maize.
However, when two inbred lines of different copy number are crossed, the Fl progeny do not exhibit the
mean of the parental types. Instead wefindrapid amplification or loss of specific sequences. We have further
evidence that tissue culture and other forms of stress can modulate the copy number of repetitive sequences
even in the absence of a genetic cross. We have also characterized the behavior of a transposable sequence
called Mutator which undergoes alterations in its capacity to transpose within one generation or even within
a single plant; the transposition of Mutator is affected by its own copy number which can readily change.
We believe that these rapid genomic changes detected in maize represent part of the plant's capacity to
modulate its genome in response to a changing environment. We propose that the plant can create new
genotypes, with new phenotypes, throughout its life span and that selection for the best-adapted regions of
the plant body then occurs. Thus, we would predict that long-lived species could produce genetically
different gametes on different branches of the plant. Furthermore, the variation in growth characteristics
seen on different branches of a plant could reflect genotypic differences. These hypotheses and data
supporting them will be discussed in the context of the developmental plasticity of plants in comparison to
animals.