lier processes and thus be essential for viability. Re- 
quiring a homozygous mutant fly to survive to adult- 
hood, where the presence or absence of an R7 cell 
can be assessed, could therefore prevent the identi- 
fication of mutations in genes that also act in signal 
transduction pathways essential for earlier develop- 
mental stages. To get around this problem. Dr. Mi- 
chael Simon and others in Dr. Rubin's group created 
a situation in which Sevenless kinase activity was 
limiting and then screened for mutations that al- 
tered the strength of Sevenless signaling. They first 
carried out a systematic genetic screen for mutations 
that decrease the effectiveness of signaling by Se- 
venless by looking for dominant enhancers of seven- 
less mutations. By adjusting the temperature at 
which flies carrying a temperature-sensitive allele 
of sevenless were grown, they could adjust Seven- 
less kinase activity to a level barely above the thresh- 
old necessary for R7 cell formation. Small reduc- 
tions in the abundance or activity of other elements 
of the pathway might then be expected to lower 
signal strength sufficiently to cause a sevenless phe- 
notype in flies grown at this threshold temperature. 
This sensitivity allowed them to identify genes en- 
coding putative downstream elements of the path- 
way by screening for genes in which inactivation of 
only one copy of the gene, which would be ex- 
pected to reduce the level of gene product by half, 
resulted in the absence of the R7 cell. Since the 
other copy of the gene remained functional, they 
were able to identify these loci even though their 
functions were essential for viability. Seven genes 
were identified in this screen. The products of four 
of the seven loci also appear to be involved in trans- 
duction of signals from another tyrosine kinase re- 
ceptor, the Drosophila homologue of the epider- 
mal growth factor (EGF) receptor. 
Ras Plays an Essential Role in the Sevenless 
Signal Transduction Pathway 
One of the loci identified in these genetic screens 
corresponds to the Rasl gene. Several previous stud- 
ies had suggested a role for Ras activity in tyrosine 
kinase function. The fact that a twofold decrease in 
the level of Rasl protein can give a dramatic efi'ect 
on signaling by Sevenless suggests that Ras also plays 
a critical role in the Sevenless signaling pathway. To 
ask whether Rasl activation alone is sufficient for 
Sevenless-mediated signaling, Dr. Mark Fortini used 
sevenless gene regulatory sequences to express dom- 
inant activating Rasl alleles in those cells of the 
developing eye that normally express sevenless. 
Constitutive activation of the Sevenless receptor has 
been shown by others to result in transformation of 
cone cell precursors into supernumerary R7 cells. 
Identical results were obtained with activated Rasl 
under sevenless gene control, suggesting that Rasl 
activation may be the primary consequence of li- 
gand-induced Sevenless signaling during R7 cell de- 
termination. 
Regulation of Ras Activity — Possible Links 
Between Sevenless and Ras 
The activity of Ras proteins is regulated by bound 
guanine nucleotides: the GTP-bound state is active, 
while the GDP-bound state is inactive. The ratio of 
GTP:Ras to GDP:Ras is determined by two antagonis- 
tic reactions. An active GTP:Ras molecule is inacti- 
vated by the intrinsic GTPase activity of the Ras pro- 
tein, a process that is greatly stimulated by RasGAP. 
An inactive GDP:Ras molecule is activated by the 
exchange of the bound GDP molecule for a GTP 
molecule, a reaction that is increased by guanine 
nucleotide exchange proteins. 
One of the four loci, Sos, identified by Dr. Simon 
and his co-workers as decreasing signal transduction 
initiated by either the Sevenless or EGF receptors, 
encodes a protein with sequence similarity to 
known guanine nucleotide exchange proteins, sug- 
gesting a model in which the activation of the Se- 
venless kinase by ligand binding would stimulate 
the activity of the Sos protein. The activated Sos pro- 
tein would then stimulate Rasl protein activity 
by promoting the conversion of GDP:Rasl to 
GTP:Rasl. Moreover, in a genetic screen for muta- 
tions that increase signal transduction, Drs. Graeme 
Mardon and Ulrike Gaul isolated mutations in a 
gene, Gapl, that encodes a protein with sequence 
similarity to mammalian RasGAP. Their genetic anal- 
ysis indicates that Gapl acts as a negative regulator 
of R7 cell determination, because loss of Gapl* ac- 
tivity leads to the formation of supernumerary R7 
cells, and reduction in Gapl* activity increases the 
eff"ectiveness of signaling by Sevenless. 
Biochemical studies will be required to deter- 
mine whether Sos or Gap 1 protein activity is actu- 
ally regulated by Sevenless activity. The available 
data do, however, permit the conclusion that the 
level of either Sos or Gapl protein activity can be a 
limiting step in the decision by the presumptive R7 
cell to become an R7 cell. The expression of Gapl is 
spatially highly restricted, whereas that of Rasl is 
not. The cell-type-specific expression and function 
of Gapl is in contrast to the widespread expression 
and function of Ras 1 and therefore suggests that dis- 
tinct GTPase-activating proteins may regulate Rasl 
in diff^erent developmental pathways. 
Dr. Rubin is also John D. MacArthur Professor 
of Genetics at the University of California, Berke- 
430 
