FIG. 1. Expression of Dlx3 and Msx1 at late gastrula. Whole, fixed albino embryos were hybridized in situ with probes for Dlx3 (A, C, E,
G, I) or Msx1 (B, D, F, H, J). At stage 12 (A–F) the posterior boundary of Dlx3 expression was significantly closer to the blastopore than for
Msx1. (A, B) Dorsal view (anterior toward top), (C, D) anterior view (dorsal toward top), (E, F) lateral view (dorsal to right). Double
hybridizations in situ were performed with Xag1 (brackets, G, H) and BF-1 (arrowheads, I, J). In the extreme anterior neural plate region,
both Xag1 and BF-1 transcripts were contained within the Dlx3 expression domain (G, I). In contrast, Xag1 and BF-1 expression was
posterior to that of Msx1 (H, J). The dashed line in I corresponds to the plane of section shown in Fig. 2.
FIG. 2. Dlx3 and BF-1 are expressed in different strata of the anterior neurectoderm. An embryo which had undergone double in situ
hybridization with Dlx3 (purple) and BF-1 (turquoise) was cryosectioned sagittally, as indicated by the dotted line in Fig. 1I. Arrows
delineate the deep boundary of the anterior neural fold (asterisk). Dlx3 is expressed in the superficial cells, while BF-1 is expressed in the
deep ectoderm. B is a magnification of the box shown in A.
FIG. 3. Differential sensitivity of Dlx3 and Msx1 expression to
inhibition of BMP signaling. Animal caps were isolated from
embryos injected with 500, 1000, 2000, or 4000 pg of RNA
encoding tBR, cultured until siblings reached stage 11, and analyzed
for Dlx3 and Msx1 expression by Northern blotting. Dlx3
expression persists at significantly higher levels of tBR than does
that of Msx1. (A) Graph summarizing densitometric analysis of
multiple exposures of X-ray film, normalized to linear film response
range. (B) Sample exposure.
FIG. 4. Differential sensitivity of Dlx3 and Msx1 expression to
inhibition of protein synthesis. Animal caps were isolated at stage
7 and cultured until stage 11 (midgastrula) in 0.33 MMR (Control)
or 0.33 MMR 1 5 mg/ml cycloheximide (CHX) and then analyzed
by RNA blotting for Dlx3 and Msx1 expression. Dlx3 RNA
accumulation was substantially blocked by this procedure, while
Msx1 expression was superinduced. w11, whole stage 11.
FIG. 5. Gene expression in animal caps injected with RNAs
encoding tBR, Msx1, and Dlx3. Animal caps were removed from
embryos injected with 2 ng tBR, 2 ng tBR1120 pg Dlx3, 2 ng
tBR1120 pg Msx1 RNA, or 2 ng tBR1120 pg Dlx31120 pg Msx1 or
from uninjected embryos, cultured until sibling embryos reached
late neurula (stage 18; W18), and processed for RNA extraction.
Northern blot analysis revealed that Dlx3 was permissive for
expression of the anterior neural marker Otx2 and the cement
gland markers CG7 and Xag1, while Msx1 inhibited expression of
these markers. Dlx3 suppressed expression of the panneural marker
gene Nrp1, the “prepattern” gene Zic3, and the anterior neural fold
gene BF1. Msx1 also inhibited Zic3 and BF1, but appeared to have
less effect on Nrp1. The epidermal keratin gene XK81 was robustly
expressed in uninjected caps and suppressed by tBR injection.
Neither Dlx3 nor Msx1, nor a combination of both factors (last
lane), increased XK81 expression significantly. Absence of signal
with Xbra verified that the animal caps were not contaminated
with mesoderm, and the accumulation of the ribosomal protein
RNA L1 was used as a control for explant viability. At the bottom
is a photograph of the ethidium bromide staining of the 18S region
from a typical gel.
FIG. 6. Dlx3 functions upstream of Zic3. Animal caps were isolated
from embryos injected with 2 ng tBR, 2 ng tBR1120 pg Dlx3, or 2 ng
tBR1120 pg Dlx31100 pg Zic3 RNA and from uninjected embryos,
cultured until sibling embryos reached late neurula (W18; stage 18),
and analyzed by Northern blot (A). Expression of the panneural
marker Nrp1 was induced by tBR, inhibited by Dlx3, and restored by
Zic3, indicating that Zic3 functions downstream of the negative
regulatory step imposed by Dlx3. Ethidium bromide staining of the
18S rRNA region is shown as a gel loading control. Absence of
hybridization to the Xbra probe indicated that the animal caps were
not contaminated with mesoderm. (B) Histogram representing densitometry
of Nrp1 hybridization signals.
FIG. 7. Disruption of anterior neural gene expression in vivo by Dlx3. Embryos were injected in a single dorsal animal blastomere at the
8- to 16-cell stage with a mixture of RNA encoding Dlx3 (40–60 pg) and b-galactosidase (300 pg). At stage 17 embryos were fixed and stained
briefly with salmon-gal to identify correctly targeted embryos, which were then hybridized in situ with probe for BF-1 (A), Zic3 (B), or Xag1
(C). The salmon-gal staining is pinkish-red, and the hybridization staining is purple. The co-injected side is to the left, and the uninjected
right side serves as a control. Dlx3 injection repressed expression of BF-1 and Zic3, but had no discernable effect on Xag1.
FIG. 8. Model of ectodermal patterning by BMP-regulated expression of Dlx3 and Msx1. A sagittal section of a midgastrula embryo is
depicted to illustrate the spatial relationships occurring along the dorsal–ventral neuraxis during gastrulation. Under the influence of the
BMP signaling (purple), Msx1 (magenta) is expressed ventrally from a point near the blastopore lip to an anterior limit relatively distant
from the leading edge of organizer tissue (yellow). Dlx3 (green) is also ventrally expressed but the domain extends from a point farther from
the blastopore lip than Msx1 and extends more posteriorly, into a region with relatively greater attenuation of the BMP signal. In the
extreme anterior neural plate, the ectoderm is divided into three domains: Msx11/Dlx31, which forms epidermis, Msx12/Dlx1, giving
rise predominantly to the cement gland (tan), and Msx12/Dlx32, which is initially specified as anterior prospective central nervous system
tissue (light blue). Thus, differential activation of Dlx3 and Msx1 by intermediate levels of BMP activity results in patterning of the
anteriormost dorsal ectoderm. Dorsal is to the right and anterior is up in this diagram.