Signal Transduction and the Specification 
of Cell Fates 
Norbert Perrimon, Ph.D. — Assistant Investigator 
Dr. Perrimon is also Assistant Professor of Genetics at Harvard Medical School. Of French nationality, he 
was educated at the University of Paris VI, where he majored in biochemistry. His thesis, with Madeleine 
Gans as advisor, was on Drosophila genetics. He moved to Case Western Reserve University as a postdoc- 
toral research fellow with Anthony Mahowald. He became a Lucille P. Markey Scholar in Biomedical 
Sciences while in Cleveland. He then assumed his present position at Harvard Medical School. 
CONSIDERING the number of cells in an or- 
ganism and the seemingly infinite possibili- 
ties for developmental errors, it is amazing that 
individuals retain their characteristic shapes. 
Cells of specific tissues interact at multiple stages 
of development, and their communication is vital 
to the establishment of patterns. Unraveling the 
mysteries of the intercellular communication 
processes is the interest of our laboratory. 
We are studying these processes in Drosoph- 
ila, since mutations are easy to generate, known 
mutant phenotypes are plentiful, and techniques 
for cloning and transferring genes are well estab- 
lished. Our goal is to identify and characterize, 
using genetic and molecular techniques, the com- 
ponents of specific signal transduction pathways 
implicated in the control of pattern formation. 
Cell fate along the anterior-posterior egg axis is 
determined by three groups of maternally ex- 
pressed genes: those of the anterior, posterior, 
and terminal class systems. We are investigating 
the terminal class system, which controls cell fate 
at both ends of the Drosophila embryo. In em- 
bryos lacking maternal terminal class functions, 
both head and tail structures are deleted. 
The activities generated by the terminal class 
system are believed to be propagated through a 
signal transduction mechanism that involves a 
phosphorylation cascade. The current model is 
that the putative transmembrane tyrosine kinase 
receptor, encoded by the gene torso, is activated 
only at the egg termini. This localized activation 
of torso then triggers a phosphorylation cascade 
that culminates in transcription of the zygotic ter- 
minal gap genes tailless and huckebein, which 
are known to encode transcription factors. 
Our molecular and genetic analysis of the ter- 
minal class system focuses on the proteins in- 
volved in transducing the signal from torso's 
membrane-bound kinase to the nucleus, and thus 
far we have characterized two genes, l(l)pole 
hole and l( 1 ) corkscrew. We have shown by ge- 
netic epistasis experiments that both act down- 
stream of the torso protein activity. The molec- 
ular characterization of l(l)pole hole has 
strengthened the current model for transfer of the 
torso signal, since the 1(1 )pole hole gene prod- 
uct is the homologue of the mammalian Raf-1 
proto-oncogene, which encodes a serine/thre- 
onine kinase. Molecular characterization of 
l( 1 Jcorkscrew, currently in progress, will pro- 
vide more insight into this signal transduction 
pathway. 
To identify other molecules involved in the 
1(1 )pole hole signal pathway, we have under- 
taken two genetic approaches to screen for sec- 
ond-site suppressors and enhancers. In the first 
type of screen, we have used a mutation in the 
1(1 )pole hole gene that reduces its level of activ- 
ity. Progeny with this gene are not viable. They 
can live, however, in the presence of a second- 
site suppressor. We have conducted extensive ge- 
netic screens to search for these suppressors and 
have recovered a number of them. Presumably, 
these suppressors identify proteins involved in 
the 1(1 )pole hole signal transduction pathway 
that are able to increase or bypass the 1(1 )pole 
hole activity. Future work will encompass a de- 
tailed characterization of these suppressors. 
In the second type of screen, we will search for 
second-site modifiers that affect the activity level 
of an l(l)pole hole gain-of-function mutation. 
Since such hyperactive gain-of-function l( 1 )pole 
hole mutations have not been recovered by con- 
ventional genetic techniques (presumably they 
are lethal to the fly), we decided to "switch on" 
an activated form of l( 1 )pole hole in nonvital or- 
gans of the adult fly to generate a visible domi- 
nant phenotype. Using the activation properties 
of the yeast GAL4 protein, which activates only 
those genes bearing a GAL4-binding site within 
their promoters, we were able to express a hyper- 
active 1(1 )pole hole-modified protein in the 
fly's eye imaginal disc. We are using the resultant 
dominant eye phenotype to screen for mutations 
that suppress the phenotype. Presumably the 
suppressors will identify proteins that can reduce 
the level of l( 1 )pole hole activity. 
We believe that such experiments will ulti- 
mately permit genetic isolation of more compo- 
nents required in the l(l)pole hole signal trans- 
duction pathway. Although our primary objective 
is to understand the role these molecules play in 
353 
