Development of the Drosophila Peripheral Nervous System 
perature-sensitive diphtheria toxin, DT-Ats, that 
we have developed in yeast and tested in Dro- 
sophila eye cells. 
Sequence analysis of the different couch po- 
tato cDNAs indicates that the gene may have pe- 
culiar features. The three cDNAs share a long 
open reading frame that conforms with the Dro- 
sophila codon usage, but we have been unable to 
define an initiation codon. Translation of the 
open reading frame indicates that the gene may 
encode a transmembrane protein that shows no 
homology to any of those published in the data 
banks. 
In summary, we believe that couch potato may 
play an important role in the development of the 
PNS, as its gene product may help guide the axons 
of the PNS neurons to their appropriate target. 
The role of the gene in adult thoracic muscle de- 
velopment remains elusive, but many phenotypic 
similarities have been observed between couch 
potato and stripe, which maps close to the 
former gene. The stripe gene also affects thoracic 
muscle development and is thought to be re- 
quired in the nervous system. 
The second gene we are studying, named A3 7, 
is expressed early in the development of the PNS. 
It maps at cytological band 80A and, like couch 
potato, it was first detected in an enhancer 
screen. So far, only a single insertion has been 
identified, to our knowledge. Flies homozygous 
for this particular enhancer detector insertion are 
homozygous viable. A mutagenesis experiment 
allowed us to recover 10 embryonic recessive 
lethal mutations that fail to complement one an- 
other. The phenorype of the mutant embryos is 
being characterized, and preliminary data sug- 
gest that some cells of the embryonic PNS are 
lacking, although the gene is expressed in almost 
all PNS cells, including the support cells. The 
gene is apparently encoded in a 4.6-kb transcript, 
which is only expressed transiently in the PNS of 
developing embryos (5-10 h). Later in embry- 
onic development (14-16 h), the transcript is 
also observed in most embryonic muscles. Prelim- 
inary sequencing data show that the cDNAs that 
we have isolated so far contain no open reading 
frames of significant length. We are presently try- 
ing to define the phenotype of the mutants more 
accurately and continuing to characterize the 
gene at the molecular level. 
36 
