Cell movement in development
115
Modulation of growth and migration of liver epithelial cells by hormones
and extracellular matrix (EGM)
E. Bade and B. Nitzgen, Fakultdtfiir Biologie, Universitdt Konstanz, D-7750 Konstanz, West Germany
A precise regulation of liver-cell growth and migration is essential for normal development and
regeneration. In vivo control is exerted through humoral (hormonal and non-hormonal) and local (cell-cell
and cell-matrix) interactions.
To study some of these control mechanisms, we used a rat-liver epithelial cell line that, at saturation,
forms a cobblestone-type lawn ('density inhibition'). The cells can be maintained in serum-free medium.
They synthesize and secrete the ECM glycoproteins fibronectin and laminin
and organize these proteins into
typical extracellular networks. The cells possess receptors for EGF (105/cell) and respond to this factor and
to insulin with growth and changes in morphology and behaviour. At optimum concentrations of insulin
(2 /Ltg/ml) plus EGF (10 ng/ml) the cells are also stimulated to migrate. Migrating cells deposit fibronectin
and laminin onto the substratum, forming well-defined ECM migration tracks. The cells can migrate at least
100 /urn per day on the plastic substratum. Both growth and migration can be modulated by components of
the extracellular matrix. Fibronectin, laminin, and collagen IV improve the survival of cells in nonsupplemented, serum-free medium, but only fibronectin inhibits migration. As a method for the study
of cell migration, analysis of ECM migration tracks in serum-free medium permits quantitative conclusions
and is not influenced by unrelated phenomena as with the study of phagokinetic tracks (A. Biihler 1977).
With this method, the liver epithelial cells here used die after incorporation of saturating amounts of
colloidal gold.
Effect of substrate and the lymphokine MIF on macrophage activity
/. M. Boswell and A. P. Swan, Department of Zoology, University of Leicester, University Road, Leicester
Mature macrophages are found concentrated in the kidney, brain, lung and liver. Populations are also
found wandering on and within the connective tissues surrounding body cavities. Macrophages are
important secretory and immunoregulatory cells and they have microbicidal and tumoricidal functions.
Resident macrophages were harvested from the peritoneal cavity of untreated mice. Mice previously given
an intraperitoneal injection of potato starch to induce a mild inflammatory response yield stimulated
macrophages which perform functions at an enhanced level compared to resident macrophages. Both these
populations of cells can be activated by specific stimuli to show enhanced levels of particular types of
behaviour.
Macrophages adhere rapidly irrespective of substrate. Initial 'Fast' spreading is inhibited by the
lymphokine MIF, (macrophage migration inhibition factor), a 60K dalton glycoprotein, and is also
dependent on the substrate provided. Macrophages spread much more rapidly and extensively on glass than
on collagen matrices. Macrophage locomotion and morphology is also affected by substrate.
Macrophages invade 3-D collagen matrices in vitro, sending large processes down into the collagen then
drawing the rounded cell body into the matrix. Stimulated macrophages were found to invade more rapidly
than resident macrophages and the addition of MIF reduces invasion of both these populations.
116
Cell movement in development
Expression of a cholinergic system during morphogenesis
Ulrich Drews, Heinrich Schmidt, Gunter Oettling and Promote Vanittanakom, Anatomisches Institut,
University of Tubingen, 4700 Tubingen, West Germany
In embryonic cells, in general, a cholinergic system is temporarily expressed during phases of
morphogenesis. The system was most extensively studied by us in the chick limb bud.
Cholinesterase (ChE). In a histochemical study using polyethylene glycol embedded chick embryos we
demonstrated ChE activity in the apical ectodermal ridge (AER), the subridge mesenchyme and the central
chondrogenic core. The embryonic ChE was characterized biochemically by colorimetry and gel
electrophoresis.
Choline acetyltransferase (ChAT). In order to prove the presence of acetylcholine we measured the
acetylcholine synthesizing enzyme in the limb bud during the phase of ChE activity.
Ultrastructural localization of ChE. In the AER, the subridge mesenchyme and in chondroblasts ChE is
localized in the endoplasmic reticulum, the perinuclear cisterna and in early stages also in the Golgi-zone.
Special contact zones to the cell membrane are observed. Embryonic ChE disappears from the cells with
cellular differentiation.
Presence of acetylcholine receptors. By binding studies with the muscarinic ligand 3H-QNB we
demonstrated a specific muscarinic acetylcholine receptor in the chick limb bud.
Chlorotetracyctinefluorescence.On stimulation with muscarinic compounds the embryonic cells respond
with intracellular Ca + + mobilization. The Ca + + fluxes are measured by monitoring chlorotetracycline
fluorescence in cell suspensions. The reaction is used for pharmacological characterization of the muscarinic
receptor and for compilation of dose response curves.
Cell migration during early embryogenesis in Apis mellifera
(Hymenoptera): SEM observations
R. Fleig*, Institut fur Biologie I (Zoologie), Albertstrasse 21a, D-7800 Freiburg, West Germany
Migration of cells and of extending free edges of epithelia plays an important morphogenetic role in the
development of the honey bee embryo and its envelopes. At the onset of gastrulation, dorsolateral
blastoderm cells move loosely connected towards the dorsal midline; they keep contact with each other and
with their substrate by short extensions while the few cells migrating singly send very long and slender
filopodia from their flattened margins in all directions. Ultimately they completely coyer the
yolk-plasmalemma of the cell-free dorsal strip and form the pre-serosa as a flat epithelium. This epithelium
then disconnects from the germ band and its free edge moves over the germ band, closing ventrally to
complete the serosal envelope. The migrating cells remain tightly joined and only the cells at the advancing
margin form globular lobopodia. A short time later the cells at the lateral free margin of the germ band start
moving dorsally underneath the serosa. At first they extend very long and slender filopodia over the dorsal
yolk plasmodium. Thereafter they and the adjacent cells loose their columnar shapes and advance as a
closed epithelium. Thus they form the thin amniotic epithelium which in turn will be overgrown by the
advancing flanks of the germ band at dorsal closure. Underneath the developing amnion, the cells of the
anterior and posterior midgut rudiments migrate along the dorsal yolk surface in opposite directions; after
covering the dorsal surface they spread ventrally and join up along the ventral midline, thereby enveloping
the yolk system. Before migrating, these cells contact each other by local interdigitation of short processes.
These then disconnect and the cells appear to migrate singly and with a tendency for mutual repulsion;
thereby they send long filopodia from all around their margins. It is only after they have spread over the
whole yolk surface that they close up, loose their filopodia and thus form the midgut epithelium. All events
described here occur within a developmental period of about 10 h, and it is truly remarkable that the
different embryonic cell types maintain their distinct morphological syndromes even if performing very
similar morphogenetic tasks. Within a cell population, the syndrome may change in the course of a few
hours without intervening mitoses (pre-serosal/serosal migration).
Cell movement in development
117
Dynamics of primordial germ cells migration in gastrulating avian
embryos
Malca Ginsburg*, Zehava Rangini and Hefzibah Eyal-Giladi, Department of Zoology, Hebrew University,
Jerusalem 91904, Israel
The exact distribution of Primordial Germ Cells (PGC) in the upper versus lower layer of early avian
blastoderms (stages XIII E.G&K to 4H&H) was studied in both chick and quail.
1) The entire lower layer was removed from the rest of the blastoderm and each part was cultured
separately. Both parts were fixed when the blastodermic fragment reached the stage 7-8 H&H (2-3
somites). Each part was serially sectioned and adequately stained, after which its entire PGC population was
counted. At the time of fixation the added number of PGCs from both fragments belonging to the same
blastoderm, was found to be relatively constant according to the final developmental stage. The total
number was also equal to the number of PGCs in unoperated control embryos and in agreement with data
from the literature. However the distribution of PGCs between the two fragments varied according to the
stage at the time of operation. At stage XIII most of the PGCs were found to be associated with the
blastodermic (epiblast) fragment, whereas at stage 4H&H they were associated with the lower layer. The
results indicate that already at stage XIII the PGCs start to leave the epiblast and migrate into the lower
layer (hypoblast+entoderm). This migration terminates at stage 4H&H (full primitive streak) when the
epiblast seems to be entirely depleted of its PGC content.
2) In another experimental series performed on stage XIII E.G&K to 3H&H blastoderms, in addition to
the removal of the hypoblast, the entire area opaca plus the adjacent more centrally located marginal zone
were also removed and the remaining central part of the area pellucida was cultured. As in the former series
the number of PGCs that developed in the margin-less blastodermic fragment decreased the older the
blastoderm was at the time of the operation.
In stage XIII blastoderms, the removal of both the marginal zone and the hypoblast, interfered with
primitive streak formation while PGC amount was maximal. It was possible to conclude the PGC
development and migration, are probably independent of the existence of a primitive streak.
A comparison of cell morphology and receptor mobility of fibroblasts
moving on planar substrata and in three dimensional fibrillar matrices
Julian P. Heath*1 and Kjell-Olof Hedlund 2. XMRC Cell Biophysics Unit, Kings College, London
WC2B5RL.2Department of Zoology, Uppsala University, Sweden
Chick embryo fibroblasts were cultured on glass coverslips and in fibrillar collagen gels. We compared the
locomotory behaviour and the mobility of surface receptors on cells moving on the two different substrata
using time-lapse video, immunocytochemistry, SEM, TEM, and stereo HVEM.
Superficially, fibroblasts appear very different on 2D and 3D substrata. On glass, moving fibroblasts are
typically fan-shaped with a oroad flattened leading lamella; in gels they are arborised with several long
narrow cylindrical pseudopodia. The net speed of locomotion is two to three times greater on glass, but this
is largely because fibroblasts in gels often extend pseudopodia, at rates comparable to cells on glass, that fail
to make contact with the collagen fibrils and are completely retracted. Structurally the differences are only
ones of scale. Interestingly, fibroblasts often retain a dorso-ventral polarity in the gels and have small
lamellae at the ends of pseudopodia. They lack the large stressfibrestypical of cells on glass, but they have a
cortical sheath of microfilaments some of which are aligned in bundles. Furthermore, there are
vinculin-containing plaques at sites of cell-matrix adhesion.
On fibroblasts moving on glass, patches of surface receptors are cleared from the dorsal surface of the
lamella and capped within 15 min, but the mobility of ventral patches is restricted. Cells in gels 'cap' patched
receptors more slowly, but receptors are cleared from all parts of the surface, collecting as a collar around
the perinuclear zone.
118
Cell movement in development
Morphological, behavioural and ultrastructural studies of avian
mesencephalic neural crest cells in vitro
D. A. Johnston and P. V. Thorogood, Department of Biology, Medical and Biological Sciences Building,
Bassett Crescent East, Southampton SO93TUand Dr D. R. Garrod, C.R.C. Medical Oncology Unit, Centre
Block, Southampton General Hospital, Southampton SO9 4XY
Morphological, behavioural and ultrastructural characteristics of cells have been compared in primary
explants of quail mesencephalic neural crest - NC (an ectomesenchyme) and of quail cardiac fibroblasts and
CFB (a true mesenchyme) growing on glass or plastic in vitro.
The motile activity of isolated NC and CFB cells at 48 h in culture has been compared. Both cell types
devote the same proportion of their margins to motile activity in active sites of similar size and with the same
distribution around the cell periphery. Time lapse studies of these cells shows that both cell types undergo
random movement and the general nature of their traces is identical. Hence, by several criteria, isolated NC
and CFB cells at 48 h are indistinguishable.
In contrast, NC and CFB cells at the edge of the expanding monolayer at 48 h in culture show completely
different behaviours. CFB cells move out radially in approximately straight, parallel tracks whilst NC cells
show far less polarity, moving in any direction from radially outwards to radially inwards. This difference in
behaviour reflects a difference in organisation of the cell's motile apparatus. The orientation of the
nucleus-microtubule organizing centre (N-MTOC) axis with respect to the direction of radial outgrowth has
been analysed for both NC and CFB cells at various times in culture. The N-MTOC axis of CFB cells is
highly polarized in the direction of radial outgrowth at 24,48 and 72 h in culture. No such polarity is evident
in NC cells at 24 or 48 h but develops by 72 h in culture.
Observations of time lapse recordings of NC cells suggest that they have a very low cell-cell adhesiveness.
Their adhesiveness relative to that of the tissues they encounter in vivo has been tested. A hypothesis
concerning the role of cell-cell adhesion in NC cell migration in vivo will be presented.
Cell contraction state and motility as studied in situ within chick
blastoderms: a physiological approach to problems of morphogenesis
P. Kucera* andM.-B. Burnand, Institute of Physiology, Medical Faculty, University of Lausanne,
Switzerland
Mechanical behaviour of cells has been studied in chicken blastoderms normally developing in a
transparent 'artificial egg'. Two techniques were used: (1) recording, by means of an oscillating microneedle, of tissue viscosity changes reflecting the state of cytoskeletal elements and (2) a real time video
processing which, by subtracting the initial image of embryo from subsequent images recorded during
experiments, allows us to visualize on-line the image of changes the tissue is undergoing (e.g. opacity
variations, individual or mass cell movements). Patterns of spontaneous mechanical activity as well as
mechanical responses to electrical and chemical stimuli were studied during the period of gastrulation.
In the area pellucida, spontaneous periodical variations of tissue contraction state were recorded. In turn,
the image analysis has suggested that these oscillations (7-12/h) might be linked to saccadic movements of
cells emerging from the primitive streak into the mesoblastic layer.
The origin and distribution within the blastoderm of mechanical tension were followed during the
expansion of blastoderm. It was found that the tension is generated actively, not by the expanding edge but
by cells located in an intermediate zone of the area opaca.
Phasic variations of tissue viscosity in response to single electrical stimuli were recorded and interpreted
as contractions of embryonic cells, eventually propagating from the stimulated region to distant areas of the
embryo. It has been observed that a series of such contractions evoked by a repetitive electrical stimulation
leads to a morphological reorganization of embryonic layers.
These physiological data will be discussed with respect to epigenetic factors which might modulate the
normal early morphogenesis.
Cell movement in development
119
Guanylate cyclase from Dictyostelium discoideum: catalytic properties
and inhibition by protease inhibitors
Marie-Lise Lacombe and Michel Veron, Biochimie Cellulaire, Institut Pasteur, 75015 Paris, France
One of the first events following chemotactic stimulation of D. discoideum, is an intracellular burst of
cGMP, presumably due to a transient activation of guanylate cyclase (GC). In an attempt to understand the
mechanism of this activation, we have studied this enzyme during starvation of Dictyostelium cells.
GC activity was very unstable: when measured at 24 °C, immediately after thawing the frozen cells in
15 % glycerol, the activity was linear for only 4 min, and at 0 °C the half-life of the enzyme was 2 h. As for
GC from higher eukaryotes, Mn-GTP was a much better substrate than Mg-GTP and an excess of free metal
(1 raM Mn2+ and 5 mM Mg2+) was necessary for optimal activity. The specific activity in vegetative
amoebae was 0-1-0-2 nmol/min/mg of protein and increased 3 fold after 5 h starvation. These values are
greater than those of the most active GC from mammalian tissues. While 90 % of the activity was found in
the cytosol, Dictyostelium GC differs from the cytosolic enzyme from mammalian origin by several
properties: app Km for Mn-GTP of 2 mM (instead of 50 /uM), MW of 270,000 instead of 150,000), lack of
sensitivity to nitric oxide, protoporphyrin IX or detergents, and ATP activation by lowering the app Km for
Mn-GTP to 0-6 mM.
Addition of antipain and leupeptin, protease inhibitors specific for serine- and SH-proteases, decreased
GC activity from aggregation competent cells by 60 %. The inhibition was less pronounced (25 %) with
vegetative cells. Maximal effect of antipain was obtained at 5 /Ag/ml with half maximal inhibition at 2 /xg/ml,
a value which is in the same range than the Ki reported for known proteases. The antipain inhibition was
due to a decrease in the Vmax with no effect on the respective app Km measured either in the absence or in
the presence of 0-2 mM ATP. Antipain did not modify the cytosolic localization of GC activity. TLCK and
TPCK inhibited GC but only at concentrations greater than 1 mM, which might reflect a non-specific effect.
No inhibition was observed in the presence of PMSF, benzamidine (both at 1 mM), soybean trypsin
inhibitor (2 mg/ml) and phosphoramidon (04 mg/ml). Thus involvement of serine- or metallo-proteases is
unlikely.
These results strongly suggest that GC is present in Dictyostelium as an inactive precursor that is activated
by endogenous proteolysis. This could provide a mechanism for GC activation upon chemotactic stimulation
of D. discoideum cells.
Stimulation of neural crest cell migration in the axolotl embryo by tissue
grafts and transplanted extracellular matrix
Jan Lofberg*1, Anita Nynas-McCoy 2, Lars Jonsson i, Roberto Perris l and Hans Henning Epperlein 3.
1
Department of Zoology, Uppsala University, Uppsala, Sweden. 2Hopkins Marine Station of Stanford
University, Pacific Grove, CA, U.S.A. z Department of Anatomy, Freiburg University, Freiburg,
West Germany
We have developed an in vivo test system to find out if the onset of neural crest cell migration in the
embryonic axolotl trunk is influenced by surrounding tissues and their associated extracellular matrix
(ECM). Tissue grafts or embryonic ECM adsorbed in vivo onto ECM microcarriers, prepared from
Nuclepore filters, were placed adjacent to the premigratory crest cells and the embryos were then incubated
to a specific stage. The experiments were evaluated with light microscopy, TEM and SEM.
We found that especially grafts of the dorsal epidermis stimulated crest cell migration locally - in the
region under the graft. The microcarrier experiments showed that the subepidermal ECM alone could
initiate crest cell migration, suggesting that the ECM of the epidermal grafts was the stimulating factor.
These results indicate that the premigratory crest cells have latent migratory capability but that they need
to be triggered from the surrounding ECM to start migration. We propose that trie subepidermal ECM, as
substrate for crest cell locomotion, is a regulating factor for the onset of neural crest cell migration in the
axolotl embryo.
120
Cell movement in development
Cell-substratum spreading is impaired in monensin-treated dystrophic
human fibroblasts
/. A. Pizzey and G. E. Jones*, Biology Department, Queen Elizabeth College, London W8 7AH
We have previously shown that the monovalent ionophore monensin suppresses the spreading of freshly
dissociated normal human fibroblasts onto a glass substratum (Pizzey, Bennett & Jones, 1983). This effect is
also seen in skin fibroblasts from patients with Duchenne muscular dystrophy (DMD) when these cells are
seeded in the presence of 5xlO~7M monensin. In both cases, cell spreading is reduced within 30 min
of seeding onto glass substrata and by 100 min, both normal and DMD monensin-treated fibroblasts have
spread considerably less (P < 0-001) than untreated cells. No difference is found in the spreading ability (or
its suppression by monensin) between normal and DMD cells, as determined either by comparing cell area
(CA) frequency distributions or by parametric significance tests on In-transformed CA values (since a
population of spreading cells do not fit a Gaussian distribution). Cell spreading is further
reduced if
fibroblasts are given prolonged monensin pre-incubation (20 h at 37 °C with 5xlO~7 M monensin),
although if such-treated cells are subsequently seeded in monensin-free media, cells then spread to a greater
extent than if the ionophore is re-introduced to the seeding media. However, prolonged monensin
pre-incubation has a more pronounced effect on the spreading of DMD cells. From the back-transformed
data of monensin-preincubated DMD cells which are allowed to spread in2 monensin-free media for 100 min,
mean CA = 198 /im2, with 95 % confidence
limits of 183-215 jam , which is significantly different
(P < 0001) from normal cells (CA = 418 /xm2, 95 % confidence limits of 319-547 /xm2). Since monensin is
known to interfere with the translocation of surface-associated glycoproteins from the Golgi system to the
plasma membrane, these results are consistent with many previous reports that the cell surface or some
surface-related mechanism may be associated with the pathogenesis of DMD.
PIZZEY,
J. A.,
BENNETT,
F. A. & JONES, G. E. (1983). Nature 305, 315-317.
The role of cell migration in blastema formation during planarian
regeneration: the use chromosomal and nuclear markers
E. Said and J. Baguna, Dept. Genitica, Univ. de Barcelona, Diagonal 645, Barcelona-28, Spain
One of the tenets of Wolff and Dubois' 'neoblast theory' of planarian regeneration (Wolff, 1964) is that
blastema is mainly formed by the accumulation of undifferentiated parencnymal cells (neoblasts) that can
migrate, if needed, over long distances to the wound. That neoblasts actually migrate was shown after
partial X-ray irradiation, and total irradiation and grafting using planarian strains of different pigmentation.
To test the extent of such migration in normal regeneration, and trying to avoid the flaws of using
pigmentation as a marker, we made grafts between sexual and asexual races of Dugesia(S)mediterranea that
differ in a chromosomal marker. Also, grafts between diploid and tetraploid biotypes of Dugesia(S)polychroa, differing in nuclear size, were done. The hosts were intact or regenerating organisms, irradiated
(X-rays; 8000 rads) or not.
The results show that: 1) migration of graft cells into host tissues occur in intact organisms in all directions
at a rate of ~40 /itm/day, rate that increase up to ~75 /xm/day in irradiated hosts; 2) regeneration does not
speed up migration rate nor drives cells preferentially to the wound; 3) cells placed at ^ 500 fim from the
wound do not participate in blastema formation; and 4) migration rate within an area of 400-500 /im around
the wound increases up to 140-150 /um/day.
It is concluded that cell migration in intact planarians is a slow phenomenon mainly linked to cell division,
and that blastema formation is a local process where only cells within an area 400-500 /urn around the
wound participate, through cell proliferation and cell migration.
Cell movement in development
121
Catecholamines induced adhesiveness in primordial germ cells of the
early chick embryo
M. Sarasa Barrio, S. Climent-Peris and L. Dominguez Roeznillo, Departamento de Anatomia y
Embriologta, Facultadde Veterinaria, Miguel Servet 177, Zaragoza-13, Spain
Catecholamines were used in this study to modify the morphogenetic movements of gastrulation in an
attempt to analyze changes in the distribution pattern of primordial germ cells (PGCs). A total of 200 eggs
were used, of which 50 were kept as controls.
Injection of dopamine or noradrenaline (100 jug/0-1 ml saline/embryo) into sub-blastodermic cavity of
unincubated eggs produced thickening and/or S-shaping of the primitive streak and prevented anterior
migration of the mesoderm. PGCs were always seen in the germinal crescent area, although they were also
seen along the blastoderm when primitive streak was S-shaped!
Unlike the PGCs of the control embryos, those of the treated embryos showed severe modifications in cell
shape. After 24 h of incubation they adhered to each other either by pseudopods or by large contact
surfaces, forming chains of several cells. After 48 h of incubation, PGCs remained adherent, forming chains
and even groups, most of them inside the extraembryonic blood vessels of the crescent, with many
lamellipodial and filopodial processes. Ultrastructural analysis showed intermediate or desmosome-like
junctions between them.
These results indicate that catecholamines raise the cytoskeleton activity in PGCs, resulting in changes of
cell shape and cell surface properties, determining adhesiveness to each other, which does not occur under
normal conditions nor under treatment with plant lectins.
Does cAMP signalling regulate multicellular morphogenesis in all
cellular slime moulds?
Pauline Schaap and Mei Wang, Zoological Laboratory, Kaiserstraat 63, 2300 RA Leiden, The Netherlands
cAMP functions as chemoattractant during aggregation in some cellular slime mould species (Konijn et al.
1968) and as morphogen in all investigated species (George, 1977; Gross et al. 1981). Some species such as
Dictyostelium minutum use other chemoattractants during aggregation.
Recently evidence was presented that oscillatory cAMP signals are secreted by the organizer (tip) of
postaggregative multicellular structures of D. minutum (Schaap et al. 1984). This signalling system is
responsible for the organization of the morphogenetic movements that lead to fruiting body construction.
Oscillatory cAMP signalling also determines the polarity of the multicellular structure and is thus
responsible for the organization of pattern formation.
To investigate a possibly general involvement of cAMP signalling in the organization of morphogenesis in
all cellular slime moulds, we undertook a comparative study of the existence and chemical identity of
oscillatory signals during the development of a variety of cellular slime mould species. The stage of
development, when oscillatory signalling first becomes evident differs from one species to the other. In D.
discoideum, D. mucoroides fuscum, D. mexicanum, P. violaceum and P. pallidum the aggregation centers
shift from continuous to pulsatile emission of chemoattractant during the aggregation process. In D.
minutum pulsatile signalling starts after the completion of aggregation and slightly before the onset of
culmination. Attraction of amoebae inside the aggregate to the center of signal emission results in tip
formation. The occurrence of pulsatile signalling at an early stage of development is correlated with the
capacity of the signalling center (tip) to organize a relatively large number of cells into a single fruiting body.
Several lines of evidence indicate that cAMP is probably involved in the co-ordination of morphogenetic
movement in the multicellular stage of all investigated species.
KONIJN, T. M., BARKLEY, D. S., CHANG, Y. Y. & BONNER, J. T. (1968). Cyclic AMP: a naturally occurring
acrasin in the cellular slime molds. Amer. Nature 102, 225-233.
GEORGE, R. P. (1977). Disruption of multicellular organization in the cellular slime molds by cyclic AMP.
Cell Differ. 5, 293-300.
GROSS, J. D., TOWN, C. D., BROOKMAN, J. J., JERMYN, K. A., PEACEY, M. J. & KAY, R. R. (1981). Cell
Patterning in Dictyostelium. Phil. Trans. R. Soc. Lond. B. 295, 497-508.
SCHAAP, P., KONIJN, T. M. & VAN HAASTERT, P. J. M. (1984). cAMP pulses coordinate morphogenetic
movement during fruiting body formation of Dictyostelium minutum. Proc. Natl. Acad. Sci. USA, in
press.
122
Cell movement in development
F-actin binding proteins in different developmental stages of
Dictyostelium discoideum
M. Schleicher, G. Gerisch and G. Isenberg, Max Planck Inst. for Psychiatry and Biochemistry, 8033
Martinsried, Munich, West-Germany
In addition to the known Dictyostelium F-actin binding proteins seyerin (40 kd), ar-actinin (95 kd) and the
120 kd and 30 kd gelating factors, several new proteins were purified and analyzed for their ability to
influence the viscosity of actin gels. Among those are a 17 kd protein, obtained from extracts of plasma
membrane fractions, a 54 kd gelation factor purified from soluble extracts and two F-actin capping proteins,
one having a molecular weight of 100 kd and a second one which is composed of 32 kd and 34 kd
polypeptides in a 1:1 molar ratio. This latter protein decreases the viscosity of F-actin in a calcium
independent manner, as measured by falling ball viscometry. Using Triton extracted cytoskeletons, it was
shown that no severing activity is associated with this protein. However, actin filaments are capped at their
fast growing end as demonstrated in the electron microscope. Monoclonal antibodies raised against several
of these proteins are used to investigate the physiological functions during different stages of development.
We found that severin, ar-actinin and the F-actin capping protein are present in vegetative as well as
aggregation competent cells. These data suggest that Dictyostelium as amoebae as well as aggregating cells
require a similar or even the same set of actin filament crosslinking, severing and capping proteins to
perform regular motility.
Human placental tissue releases a potent chemoattractant for vascular
endothelial cells
Heikki Seppd and Kalman BUki, Department of Anatomy University of Oulu, Kajaanintie 52 A, SF-90220
Oulu 22, Finland
Formation of new blood vessels occurs by sprouting from previously existing microvasculature. This
process is called angiogenesis. Directed migration is an essential feature in the angiogenic response and
probably involves chemotaxis of vascular endothelial cells towards chemical signals released from the target
tissue (1,2,3).
We have used the Boyden chamber assay to detect chemotactic activity for vascular endothelial cells in
extracts of human placenta. Slices of placental tissue were found to release endothelial cell chemoattractant
activity into their incubation medium. The main activity eluted as a large molecular weight complex in gel
filtration, and can be dissociated with TCA-ethanol treatment. To isolate the activity from placenta they
were treated with TCA and extracted in ethanol. The activity present in ethanol elutes from Sephadex
LH-20 at a volume corresponding to a Mr of 200-1000. The attractant has specificity for vascular endothelial
cells as it does not attract fibroblasts or leukocytes.
We speculate that the major stromal macrophage-like cells of the villi control the vascularity of the
placenta by releasing this chemical signal.
1) GLASER, D'AMORE, SEPPA et al. (1980). Nature 288, 483-484.
2) SEPPA, SEPPA, LIOTTA et al (1981). /. Dent Res. 60A, 319.
3) SEPPA, SEPPA, LIOTTA et al. (1983). Inv. Metast. 3, 139-150.
Cell movement in development
Cephalic neural crest migration in the rat embryo
5. S. Tan* and G. M. Morriss-Kay, Department of Human Anatomy, South Parks Road, Oxford
123
OX13QX
The head region of rat embryos was investigated by SEM following removal of the surface ectoderm with
adhesive tape. Observations were made in embryos from 5-somite to 11-somite stages of development, in
order to determine: (1) the sequence of emigration of neural crest cells from the different regions of the
future brain; (2) the appearance of crest cells before, during, and after their conversion from an epithelial to
a mesenchymal form; (3) the migratory pathways.
Emigration begins in the mesencephalon, followed by the metencephalon; crest cells from these two
regions migrate into the first visceral arch. Subsequently cells emigrate from the myelencephalon, but not in
a rostro-caudal sequence. At the time of crest cell emigration, the neural fold morphology varies from a flat,
widely open plate (mesencephalon) to a closed tube (lower myelencephalon). Thus the timing of emigration
is related neither to age nor to morphology of the neural epithelium. We propose that this lack of correlation
may be due to crest cell migration beginning by caudo-rostral migration within the neural epithelium,
evidence for which is available from other studies.
The pathways of migration appear to be determined chiefly by the availability of space between adjacent
structures (neural ectoderm, surface ectoderm, and paraxial mesoderm), although cell-free spaces were not
observed to form ahead of the migrating cells. Crest cells in migration were closely associated with
extracellular matrix material, but there was no evidence of matrix fibril alignment.
In vitro evidence for inhibition of cell motility in the endophyllic
crescent of the chick blastoderm
J. Van Hoof1, F. Harrisson 1, Ch. Vanroelen \ L. Andries x and L. Vakaet 2. ^Department of Anatomy and
Embryology, State University of Antwerp, Department of Anatomy and Embryology, B-2020 Antwerp,
Belgium. 2State University of Ghent, Belgium
The mechanism of cell migration during embryonic development is hardly understood. In the chick
embryo, extracellular fibronectin-rich fibrils have been described at early stages of development at the base
of the epiblast, along the anterior border of the area pellucida, a region referred to as the endophyllic- or
germinal crescent (Critchley et al. 1979; Wakely and England, 1979). These authors suggest that these fibrils
may guide mesoblast- and primordial germ cells. Recent in vivo observations, however, suggest that they
more likely act as a spatial barrier which inhibits cell spreading (L. Andries et al. submitted). The present
study describes an in vitro experimental system in which the spreading capability of the graft on or nearby
the fibrils is tested. Grafts of the primitive streak (nodi anterior or posterior, or mid parts) were explanted
(1) in the immediate vicinity of the fibrillar zone or (2) on the band of fibrils, on the ventral side of the
epiblast of glutaraldehyde-fixed blastoderms deprived of their hypoblast after fixation. Time-lapse
photomicrographic observations of these tissues made in culture show an excentric outgrowth and spreading
of the graft explanted adjacent to the fibrils, in a direction that is perpendicular to and away from the fibrils.
Spreading of the graft was not observed in the fibrillar zone. The exact relation between the graft and the
band of fibrils was verified with SEM. It was observed that the graft firmly attached to the substrate by the
extensive presence of filopodia and lamellipodia extending from the graft cells to the epiblast. However, in
the fibril region, the same graft had a different appearance. The cells tended to bridge the fibril region, thus
lacking contact with a substrate. The graft cells in this region had only few lamellipodia. Only in this part of
the graft, discrete ruffles were present on the graft cells. These results which are concomitant with the in
vivo observations mentioned above are largely in contradiction, at least for the types of cells explanted in
our system, with the views of the above mentioned authors who regard thesefibrilsas an adequate substrate
for contact guidance of cells in the direction of the fibrils. Whether this barrier for migration of primitive
streak cells has only mechanical reasons, and so directly related to the presence of these fibrils, or whether it
is related to the presence of particular chemical compounds, remains to be investigated.
CRITCHLEY, D. R., ENGLAND, M. A., WAKELEY, J. & HYNES, R. O. (1979). Nature 280, 498-500.
WAKELY, J. & ENGLAND, M. A. (1979). Proc. R. Soc. Lond. B 206, 329-352.
124
Cell movement in development
J. Vasiliev* (Moscow)
No abstract for publication
Contact guidance in vivo — an analysis of mesenchyme cell movement in
the teleost fin bud
A. Wood* and P. Thorogood, Department of Biology, University of Southampton, Southampton SO9 3TU
Direct observation of mesenchyme cell movement in vivo has been achieved in a system containing a
structurally ordered extracellular matrix comparable to the substrata eliciting 'contact guidance' behaviour
in cultured tissue cells. The pectoral fin bud of the developing teleost initially displays an apical ectodermal
ridge, but this ridge is soon replaced by an apical fin fold in which the ectodermal epithelium becomes folded
to enclose an extracellular 'space' between the apposed basal surfaces. Collagenous fibrils or actinotrichia,
up to 2 fim in diameter, are laid down in a highly ordered arrangement in two (dorsal and ventral) arrays;
one subjacent to the dorsal epithelium and the other beneath the ventral epithelium (1). Mesenchyme cells
migrating distally from the base of the fin bud, into this space, encounter the collagen fibrils and move
between them, apparently using the fibrils as a substratum. In the killifish Aphyosemion scheeli, the entire
fin structure is transparent and using this system we have investigated the in vivo migration of the
mesenchyme cells within organ cultured fin buds.
Using Nomarski differential interference contrast microscopy and time-lapse video recording we have
found that the number of cellular processes per cell increases significantly during the onset of migration and
that these processes can be classified according to their diameters. Processes of diameter greater than 2 fim
are not usually aligned along actinotrichia and arise at any aspect of the cell body. In contrast, processes
with diameters of less than 2 /Ltm appear to be confined to the distal aspects of the migrating cells and show
an increasing tendency to become aligned as development progresses. Time-lapse video recordings reveal
such aligned processes move faster than non-aligned processes. The period of the fastest rate of cell
translocation correlates with maximum process alignment along actinotrichia (2).
Thin (1 /u,m) plastic sections reveal that, generally aligned processes are in close association with the
surface of the actinotrichial fibrils and not the spaces between them. Using computer-based image analysis
techniques and transmission electron microscopy we have made a morphometric survey of actinotrichial
diameters and the inter-actinotrichial spaces at three levels proximal to the distal edge of the fin fold and at
selected stages of fin development. These data together with details of the ultrastructure of the cell processes
and their contact relationships with the actinotrichia will be presented and discussed in the general context
of cell movement and contact guidance phenomena.
1) WOOD, A. T. (1982). Anat. Rec. 204, 349-356.
2) WOOD, A. & THOROGOOD, P. (1984). /. Cell Sci. 66, 205-222.