|
FIG. 1. Alignment of the Xstbm, mouse Ltap, human KIAA1215 and Drosophila Strabismus amino acid sequences. The conserved amino
acids are represented by asterisks (*). The PDZ-binding domains are shaded gray.
|
|
FIG. 2. Temporal and spatial expression of Xstbm. (A) Quantitative
RT-PCR was performed by using 1 g of total RNA extracted
from Xenopus embryos at different stages. “U” indicates the
unfertilized eggs, and numbers indicate the developmental stages
according to Nieuwkoop and Faber (1967). Maternal transcripts
decreased gradually until the gastrula stage (stage 10), and zygotic
expression increased thereafter in the gastrula (stage 12) and
neurula (stage 15–20). (B–O) Localization of Xstbm transcripts by
whole-mount in situ hybridization. At the gastrula stage, Xstbm is
expressed in the dorsal region (B, dorsal view, stage 10; C, lateral
view, stage 12). Xstbm is expressed in the neural plate of the early neurula (D, stage 14) and late neurula (E, stage 17), and in the neural
tube thereafter in the tailbud stages (F, stage 20; G, stage 25, dorsal
view; H, stage 25, lateral view; I, stage 30, lateral view). At tailbud
stage, Xstbm is also expressed in prenephritic region (I, yellow
pointer). Sections of the gastrula (B, C) show low expression at the
dorsal lip (red pointer) and higher expression in the posterior
mesodermal–neural region (yellow pointer) above the blastoporal
lip (J, stage 10). The late gastrula shows high expression in the
prospective forebrain (K, stage 12, yellow pointer) and low expression
in the anterior mesodermal region (K, red pointer). The
superficial epithelial layer of the involuting mesodermal region
showed little or no expression at stage 10 (J, white pointer) and
stage 12 (K, white pointer). The posterior mesodermal regions of a
dorsal sandwich explant showed expression at the late neurula
stage (L, yellow pointer), but expression declined anteriorly (L, red
pointer). Transverse sections of a tailbud stage embryo in (I) showed
the expression of Xstbm increased progressively from the anterior
(M, yellow pointer) to the middle (N, yellow pointer) and posterior
notochord (O, yellow pointer).
|
|
FIG. 3. Overexpression of Xstbm interfered with posterior neural fold fusion and closure of the neural folds to form the neural tube. (A, C, E) Neurula-stage embryos (stage 20). (B, D, F) Tailbud-stage embryos (stage 30). (A, B) Normal embryos. (C, D) Under moderate doses of Xstbm, neural fold fusion and posterior neural tube closure were impaired, and the dorsal axial and paraxial tissues and the neural plate did not converge and extend well. (E, F) Under larger doses, neural fold fusion did not occur and the dorsal axial tissues were very short in the anterior–posterior axis. Bars, 0.5 mm.
|
|
FIG. 4. Effects on phenotype by increasing doses of Xstbm expression.
Blue bars represent ratio of embryos which were severely
blocked neural fold closure as shown in Figs. 3E and 3F. Green bars
represent ratio of the embryos which were impaired neural fold
closure as shown in Figs. 3C and 3D. Yellow bars represent normal
embryos. The proportion of severely affected embryos increased
with the dose.
|
|
FIG. 5. Whole-mount RNA in situ hybridization shows expression
of neural and mesodermal marker genes in Xstbm (200
pg)-injected embryos at stage 20 (A–L). The left column shows
controls and the right column shows the corresponding Xstbminjected
embryos. The neural crest marker Xslug shows the cranial
neural crest of Xstbm-injected embryos farther from the midline
and closer to the blastopore (B), compared with the normal embryo
(A). The pan-neural marker, nrp-1, shows a very wide, short, and
unclosed neural plate typical of the Xstbm-injected embryos (D)
compared with the elongated, converged, and fusing neural tube of
normal embryo (C). The marker of prospective neurons, n- tubulin,
shows prospective neurons in a short array far from the midline
and wrapping around the blastopore of Xstbm injected embryo (F),
compared with the elongate, medial array in the normal embryo (E).
Numbers indicate corresponding components of the expression
pattern in (E) and (F). The pattern of expression of the prospective
notochord marker, chordin, is short, wide, and thick in Xstbminjected
embryos (H, dorsal; J, lateral), whereas it is elongated and
narrow in normal embryos (G, dorsal; I, lateral). The prospective
somitic mesoderm marker, MyoD, is expressed around both sides
of the unclosed blastopore in the Xstbm-injected embryo (L), whereas
it is expressed in elongate arrays on both sides of the
notochord in normal embryos (K). Dorsal explants of normal
(control) embryos showed convergence and extension of both
mesoderm and neural regions (M), whereas neither region showed
convergence and extension in Xstbm-injected embryos (N).
|
|
FIG. 6. Procedures for making open-faced explants. (A) Scattered double-labeled open-faced explants allow observation of cell behavior.
Xstbm GFP (200 200 pg) or GFP (200 pg) mRNA (green) was injected into two dorsal blastomeres of the four-cell-stage embryo (far left).
Then, Alexa594 (10 pg; red) was injected into several dorsal blastomeres of the same embryos at stage 7 (second from left). Dorsal open-faced
explants were then dissected from the injected embryos at stage 10, and any involuted endodermal and mesodermal cells were shaved off
their inner surfaces before culturing for time-lapse imaging (Fig. 7). (B) Grafts between normal and Xstbm-injected open-faced explants were
done to analyze the behavior of large and small populations of Xstbm-injected cells. Xstbm GFP (200 200 pg), GFP (200 pg) mRNA
(green), or Alexa594 (800 pg; red) was injected into two dorsal blastomeres of four-cell-stage embryo. For results in Figs. 8A and 8B, dorsal
open-faced explants were made at stage 10 (as in A), and then clumps of about 10 cells (for small populations) or hundreds of cells (for large
populations) of the mesodermal region were removed from the Xstbm GFP-injected (middle row) and GFP-injected (bottom row) explants
and wedged into Alexa594-injected explants (top row). For results shown in Fig. 9, clumps of about 10 cells (for small populations) and
hundreds of cells (for large populations) were removed from the Alexa594-injected explants (top row) and wedged into Xstbm
GFP-injected explants (middle row). (C) To analyze the effects of Xstbm on neural convergent extension for results in Figs. 8C and 8D,
|
|
FIG. 7. Frames from time-lapse movies of fluorescently labeled cells show that cells of explants made from normal embryos adopt the
bipolar (arrow), mediolaterally oriented protrusive activity typical of this region (notochord) and intercalate between one another during
convergence and extension (A), whereas the cells of explants made from Xstbm-injected embryos show unorganized protrusive activity and
did not form the bipolar cells characteristic of the prospective notochord region (B, arrow). An enlargement of the last frame is shown at
the far right. The direction of anterior–posterior axis is indicated by the dashed line, anterior at the bottom. The time elapsed is indicated
at right bottom. This figure is presented in a gray-scale version because the contrast is better than it is in the two-color version. Xstbm GFP (200 200 pg; middle row), GFP (200 pg) mRNA (green, bottom row), or Alexa594 (800 pg; red, top row) was injected into
two dorsal blastomeres of the four-cell-stage embryo (left column). Dorsal open-faced explants from Xstbm GFP- and GFP-injected
embryos were made at stage 10 as the methods in (A) (second column from left); then clumps of about 10 cells (for small populations) or
hundreds of cells (for large populations) of neural region were removed from the Xstbm GFP-injected (middle row) and GFP-injected
(bottom row) explants and wedged into the neural region of Alexa594-injected host embryos (top row) at stage 10 (middle column). After
the host embryos developed to stage 12, the dorsal neural and mesodermal tissues were removed and the superficial epithelium peeled off
the neural plate at stage 12, exposing the deep cells to time-lapse imaging (two left columns).
|
|
FIG. 8. Time-lapse movie frames show differences between the behavior of normal GFP-injected cells (green-labeled cells on left of each
frame) and the behavior of Xstbm-injected cells (green-labeled cells on right of each frame) grafted into normal, Alexa594 dextran-labeled
host explants (red background) in the dorsal mesodermal (A, B) and neural (C, D) regions. The two far right columns are high magnifications
of the last frame of the GFP-injected (left) and Xstbm GFP-injected cells (right). The explants were prepared as described in Fig. 6B for
mesodermal explants and Fig. 6C for neural explants. A large population of normal cells intercalated into the host explants and form the
bipolar cells in the prospective notochord region (green labeled group on left, A). In contrast, a large population of Xstbm-injected cells could
not intercalate into the mesodermal region of the host explant (green labeled group on right, A). However, small populations of both normal
cells (green labeled cells on left, B) and Xstbm-injected cells (green labeled cells on right, B) could intercalate among the host explant
mesodermal cells. In the neural region, large populations of normal neural cells could intercalate into the normal neural host cells (green
labeled group on left, C), whereas Xstbm-injected neural cells could not intercalate (green labeled group on right, C). However, small
populations of both normal neural cells (green labeled cells on left, D) and Xstbm GFP-injected neural cells (green labeled cells on right,
D) could intercalate into the neural region of host explants. The anterior–posterior axes in the mesodermal explants (A, B) are vertical with
anterior at the bottom at the outset (0:00) but become tilted in the course of convergence and extension (dashed lines). The
anterior–posterior axes in the neural explants (C, D) are vertical with the anterior at the top. The time elapsed is indicated at bottom right.
|
|
FIG. 9. Frames from time-lapse recording show that a large population of normal cells can intercalate between one another and converge
and extend with a surrounding host population of Xstbm-injected cells, and that at the border of the two, the Xstbm-injected (dark cells)
and normal cells (light cells) can also intercalate between one another, such that the Xstbm-injected cells (pointer) wind up among the
normal cells. At the far right are high magnifications of the last frame of the time-lapse recordings. The Xstbm-injected cells adopted the
normal bipolar shape within the large population of normal cells (pointer). In contrast, the cells of the small, normal population move and
spread out into the Xstbm-injected cells, but they did not adopt the bipolar shape, did not undergo an organized cell intercalation, and did
not contribute to convergence and extension (arrow). The elapsed time is indicated at bottom right. This figure is presented in a gray-scale
version because the contrast is better than it is in the two-color version.
|
|
FIG. 10. Diagrams show the structures of the constructs Xstbm,
Xstbm(PDZ-B), and Xstbm(TM) (A). The four prospective transmembrane
regions (TM) are indicated (107–236) and the black box
at the C terminus represents the PDZ-binding motif (518–521) (A,
top). Xstbm(PDZ-B), the PDZ-binding domain of Xstbm was
deleted (A, middle). Xstbm(TM), the TM domains of Xstbm were
deleted (A, bottom). Overexpression of increasing amounts of
Xstbm(PDZ-B) results in a dose-dependent increase in defects of
convergence and extension (B, 3 left bars). However, expression of
increasing proportions of Xstbm(PDZ-B) relative to Xstbm, results
in a dose-dependent decrease in the severity of the effect of
Xstbm on gastrulation and convergent extension (B, 3 pairs of bars,
right). Xstbm(TM) had a synergistic effect when expressed with
Xstbm (C). Xstbm(TM) shows an increasing effect on convergence
and extension and gastrulation when expressed alone (C, 3 left
bars), and acts synergistically when expressed with Xstbm (C, 3
pairs of bars on right). The blue bars represent the proportion of the
embryos that showed severely blocked neural fold closure and very short axes as illustrated in Figs. 3E and 3F. The green bars represent
the proportion of the embryos that showed impaired neural fold
closure as illustrated in Figs. 3C and 3D. The yellow bars represent
normal embryos.
|
|
FIG. 11. The effects of Xstbm(PDZ-B) on cell polarity.
Xstbm(PDZ-B) GFP (25 5 pg) was injected into several
dorsal blastomeres of a stage 7 embryo, and dorsal open-faced
explants were made at stage 10. A small, scattered population of
Xstbm(PDZ-B) GFP-expressing cells was obtained at the left
and a large, cohesive population of Xstbm(PDZ-B) GFPexpressing
cells was obtained at the right of an explant (shown in
epi-illumination, A, and in fluorescence, B). Note the elongation
and alignment of the cells at the left, whereas those at the right
remain rounded (A, B). The scattered, labeled Xstbm(PDZ-B)-
injected cells intercalated among the unlabeled normal cells (left
side) but the cohesive, labeled Xstbm(PDZ-B)-injected cells
could not intercalate with one another (right side). In this explant,
the anterior–posterior axis and axis of extension is left to
right.
|
|
FIG. 12. We propose a model of Xstbm function accounting for our results. Xstbm is localized to the plasma membrane and functions in
a complex requiring at least two Xstbm molecules with PDZ-binding domains, that together interact with a PDZ protein or proteins,
perhaps Dishevelled (see Park and Moon, 2002), to transduce a signal important in regulating polarized cell behavior (A, left diagram).
Overexpression of Xstbm would result in upregulation of this signal and repression of convergence and extension. When Xstbm(PDZ-B),
lacking the PDZ-binding region, is expressed, the dimer- or multimer-dependent PDZ-binding function of the complex fails, and no signal
or a reduced signal is transduced (B). Increased expression of Xstbm(PDZ-B) would progressively reduce the normal signal level, also
leading to failure of convergence and extension. Increased expression of Xstbm(PDZ-B) would also counter the effects of overexpression
of Xstbm, resulting in correction of convergence and extension. Expression of Xstbm(TM), lacking the transmembrane regions, would not
be localized to the membrane and thus would function inefficiently in increasing the signaling level (C). This would account for the weak
additive effect of coexpression of Xstbm(TM) and Xstbm in reducing convergence and extension.
|