Ascidian tadpole larvae have a similar dorsal tubular nervous system as ver
tebrates. The induction of brain formation from a4.2-derived (a-line) cells
requires signals from the A4.1-derived (A-line) cells. However, little is
known about the mechanism underlying the development of the larval peripher
al nervous system due to the lack of a suitable molecular marker. Gelsolin,
an actin-binding protein, is specifically expressed in epidermal sensory n
eurons (ESNs) that mainly constitute the entire peripheral nervous system o
f the ascidian young tadpoles. Here, we address the role of cell interactio
ns in the specification of ESNs using immunostaining with an anti-gelsolin
antibody. Animal half (a4.2- and b4.2-derived) embryos did not give rise to
any gelsolin-positive neurons, indicating that differentiation of ESNs req
uires signals from vegetal cells. Cell isolation experiments showed that A4
.1 blastomeres induce gelsolin-positive neurons from a-line cells but not f
rom b4.2-derived (b-line) cells. On the other hand, B4.1 blastomeres induce
gelsolin-positive neurons both from b-line cells and a-line cells. This is
in sharp contrast to the specification of brain cells which is not affecte
d by the ablation of B4.1-derived (B-line) cells. Furthermore, basic fibrob
last growth factor (bFGF) induced ESNs from the a-line cells and b-line cel
ls in the absence of vegetal cells. Their competence to form ESNs was lost
between the 110-cell stage and the neurula stage. Our results suggested tha
t the specification of the a-line cells and b-line cells into ESNs is contr
olled by distinct inducing signals from the anterior and posterior vegetal
blastomeres. ESNs in the trunk appear to be derived from the a8.26 blastome
res aligning on the edge of presumptive neural region where ascidian homolo
gue of Pax3 is expressed. These findings highlight the close similarity of
ascidian ESNs development with that of vertebrate placode and neural crest.
(C) 2001 Academic Press.