Fig. 1. Biomechanical interpretation of gastrulation
events in amphibian embryos. (AD,
E1, E2) Biomechanical �maps� of successive
stages and most important regions of
embryos. Active forces are displayed by red
arrows, while passive stresses are indicated by
blue arrows. (A) Stretching of the blastocoel
(br) roof by turgor pressure in the blastocoel.
(B) RCI in the blastocoel roof producing pressure
forces onto vegetal regions. (C) A detailed
scheme of the region framed in (B). (D) Two
alternating phases of biomechanical interactions
between the marginal zone (MZ) and the
suprablastoporal zone (SBZ). (E1, E2) Successive
stages of the involution in saggital section.
Frames (ab,c,d,e) display suggested hyper-restoration responses that correspond to the upper row frames indicated by the same capital letters (ab
corresponds to A and B together). Dotted arrows connect the active branches of preceding hyper-restoration responses with the shifts of stress values
which they produce in the neighboring parts of embryos triggering the latter�s active responses. Pre, preinvoluted; pst, postinvoluted layer.
Fig. 2. Formation of a very much diminished copy of a normal amphiblastula
from a ventral ectoderm explant of the early gastrula X. laevis embryo. (A)
Scheme of operation. (B-D) An explant immediately after extirpation, 8 and 24 h later.
From Beloussov and Petrov, 1983.
Fig. 3. Results of relaxation of tensile stresses at the surface of the late blastula
stage of X. laevis embryos. (A) An embryo with a wedge-shaped implant causing
relaxation, 1 h after operation. (B) A similar embryo, 5 h after operation. (C) Scanning
electron microscopy view of a relaxed embryo, 1 day after operation. Note two
extensive protrusions which stretch the surrounding tissue areas. (D) A saggital semithin
section of a relaxed embryo maintained for 5 h in 50% MMR solution. (E) A
similarly treated embryo maintained for the same time period in 25-fold diluted MMR,
tensions being considerably restored. (F) Fragment of the blastocoel roof of an intact
embryo of the same age as (B). (G,H) Similar fragments from (D,E) embryos,
respectively. Note the reduction of the number of cell layers in (F) as compared with
(B) and in (H) as compared with (G), indicating the renewal of RCI after incubation in
hypotonic solution. (I) A diagram summarizing the results of �stress therapy� of the
relaxed embryos. (A) Control group; (B) relaxation by a frontally oriented radial cut
through a vegetal embryo region; (C) a similar relaxation by a saggitally oriented radial
cut; (D) embryos incubated in a hypotonical (25-fold diluted) MMR after frontally
oriented cut; (E) similarly treated embryos after saggitally oriented cut; (F) frontally
oriented cut followed by stretching of the embryos. Figures in brackets give the
number of samples in the corresponding series. Vertical axis, percent of normalized
and abnormal embryos 24 h after operations. Graph 1, normal embryos; 2,
exogastrulation; 3, incomplete gastrulation; 4, other anomalies. Note a substantial
recovery after tension renewal. From Beloussov and Ermakov (2001).
Fig. 4 (Left). Relaxation-promoted formation of additional groups of bottle-shaped cells and the corresponding invaginations (shown in A,
D by dense pointers) (A) 1 h after the relaxation of the suprablastoporal zone due to its separation from the underlain tissues (large bent arrow) at
the early gastrula stage. blp is the normal blastopore. (B) A saggital section of the dorsal blastoporal lip of the normal mid-gastrula stage embryo.
Arrows show separation directions which relax pre-existed tensions. (C) A relaxed lip immediately after separation (D) 1 h later, a new invagination
pit has been formed at the site of a former dorsal blastoporal lip (pointer).
Fig. 5 (Right). Schemes of the operations for modulating σc and/or σm values (shown by dotted lines) in the marginal zone of early gastrula
embryos and expected results. (A) A transversal view of a dorso-medial cut (vertical arrow in A) providing wound opening (diverged arrows) and
hence σc relaxation, most of all in the ventral region (wavy part of a contour). Wound gap is covered by a piece of ventral ectoderm. (B) Expected cell
convergence along a ventral mid-line due to increase of inequality between a preserved σm (vertical bidirectional arrow) and relaxed σc (convex arrow).
View from the ventral side. (C) Enhancement of σc in the dorso-medial zone as a result of transversal stretching by two needles (arrows coming from
black spots). (D) Expected result; an extensive transversal cell convergence in the area of the suprablastoporal zone and latero-ventral cell movements
due to some σc relaxation in the ventral region.
Fig. 6. Inversion of cell convergence movements as a result of a dorso-medial cut of early gastrula stage X. laevis embryos. In the vertical
panels (A,B,D), left rows are optical light views and right rows luminescent views of the same embryos at the different time periods after cuts (shown
to the left). In panels (C), only luminescent views are given. In (A-C), the descendants of two ventral blastomeres at 32 cell stage are labeled, while
in panel (D), the descendants of the same stage 2 dorsal blastomeres are labeled. (A) Normal dorsalwards convergent movements of a labeled ventral
tissue. 24 h embryos from this panel are shown from the left side while all the others are displayed in the vegetal projection. (B) Inversion of the labeled
cell movements towards ventral midline as a result of a dorso-medial cut. Note that at 24 h, an abnormal tail is formed from these cells in the ventral
location. (C) If making a similar dorso-medial cut of an isolated marginal zone, no labeled cell convergence towards a ventral midline takes place. (D)
After an operation similar to that shown in panel (B), the descendents of the dorsal labeled cells move to the ventral side, that is, opposite to the normal
convergence direction. As a result, they spread along the lateral lips of the blastopore.
Fig. 7 (Left). Induction of the active lateralwards movements of the labeled descendants of dorsal cells as a result of transversal
suprablastoporal zone stretching by four needles. To the left: optical light and luminescent views just after inserting needles and before stretching
them (A), immediately after stretch (B) at 2.5 h (C), 3.5 h (D) and 7 h (E). In optical light views, needles are denoted by black spots. To the right are shown
the positions of the needles 1-4 before stretching, immediately after stretch and 2.5 h later. Note a considerable post-stretching increase of the distance
1-2 in spite of the needles being anchored to the agarose substrate. As a result, the dorsal cell material is very much shifted to the ventral, considerably
overlapping the needles� positions. This result is quite similar to that obtained by a dorso-medial cut (see Fig. 6).
Fig. 8 (Right). Morphological results of suprablastoporal zone (SBZ) transversal stretching, as compared with a similar stretch of neurula
stage embryos. (A) Scheme of SBZ stretching by two needles. (B-F) 24 h results. (B,C) Total views (on frame B, the direction of stretching is
indicated), whereas (D-F) are histological sections made in transversal embryo planes. Note a complete transversal reorientation of the embryo body
including not only the axial rudiments (notochord and somite series), but also the yolk-containing compartments. In (D) and to some extent in (F) part
of the notochord is shifted to the lateral blastoporal lip. (G) A scheme of a similar operation performed at the early neurula stage. (H) Its result, total
view. Contrary to (A-F), the larva fully preserves its initial axis, remaining bound to one of the needles by a thin unstretched tissue thread (later on
disrupted). In (H), pointers indicate the positions of the needles.
Fig. 9 (Left). Evidence of cell convergence on artificial folds and ectodermal �scrolls�. (A) A scheme of fold preparation; an ectodermal tissue
piece situated lateral to the dorsal mid-line of a gastrula stage embryo is cut off, bent as shown by a dashed line and fixed in this position by a needle.
(B,C) Protrusions formed in 10-15 h out of the bent folds (opposite to the needles). (D) If a framed region of lateral ectoderm is extirpated from the
neurula stage embryo, it rolls spontaneously into a scroll, its longitudinal axis coinciding with that of embryo (E). Soon the scroll is transformed into
a meridionally shrunk body with many transversal grooves and ribs (F,G). Times after scroll formation are shown.
Fig. 10 (Right). Residual deformations (RDs) in stretched or shrunk explants of embryonic tissues after their unloading. Vertical axes: percents
of the positive (fixed elongation) or negative (fixed shrinkage ) RDs. Initial length is taken as zero. Horizontal axes: time, minutes. (A) Protocol graphs of
a stretched explant (1) and two shrunk explants (2,3). Explant 1, stretched to 70% at 5 min time and unloaded 6 min later, gradually returned to its initial
length. Meanwhile, explant 2 which was shrunk to 50% and kept in this state for only 2 min gave - 35% RD; explant 3 shrunk for 5 min and gave -100%
RD. (B) RD after unloading 10-20% stretch in 5, 30 and 60 min after its application. (B1) Explants of the early gastrula ventral ectoderm, (B2) SBZ tissue,
(B3) explants of the lateral ectoderm from neurula stage embryos. While in 5 min after force application the samples B1 and B2 undergo 20-10%
contraction, (that is, negative RD), after 60 min stretch, B1 samples gave +20% RD, B2 samples just a slight RD and B3 samples no RD at all.
Fig. 11. Exchange of the passive stretching of the postinvoluted layer (pst)
at the early-mid gastrula stage (A) by its active extension (B). Arrows show
the directions of the separation of the pre-involuted (pre) and postinvoluted
layers. Samples were fixed in Bowen�s fluid a few seconds after separation and
sectioned saggitally. (A) An extensive contraction and curling of pst. (B) Its
active extension, overlapping the layer pre. In the both frames the anteroposterior
direction goes from up to down.
roof, stretched for an hour up to 20% of their initial length retained,
after stretch relief, just 10% of their initial length. Consequently,
the rates of normal gastrulation movements roughly satisfy the
condition of constant tension. One may conclude that the rate of
tension relaxation is the limiting parameter of gastrulation movements.
On the other hand, the lateral ectoderm explants taken from
early neurula embryos (stages 13-15) did not show RD at all after
any amount and duration of stretch (Fig. 10 B3). Hence, at these
stages the imposed tensions are not relaxed at all. One concludes
that the ability to relax tensions by cell intercalation is a stagespecific
property of gastrula stage embryos only.
Step 4: Extension of the postinvoluted cell material
EE feedback (see Beloussov and Grabovsky, this volume) is
expected to be established between preinvoluted (pre) and
postinvoluted (pst ) cell material. Namely, at the beginning of
involution pst is assumed to be passively extended by the actively
extending pre, while later on pst becomes an active component
of the mutually promoting EE feedback (see Fig. 1 E1, E2, e).
Predictions and experiments
While carefully detaching the adjacent pre and pst areas from
each other, one should expect that at the start of involution pst will
be immediately and extensively contracted (relieving the passive
tension), while at the later stages it can even overlap the overlain
pre. Just this was observed: if we made such a detachment at