GENETIC ANALYSIS OF PATHWAY AND TARGET RECOGNITION IN DROSOPHILA 
Corey S. Goodman, Ph.D., Investigator 
The molecular mechanisms that control how neu- 
ronal growth cones find and recognize their correct 
targets during development continue to be the focus 
of Dr. Goodman's laboratory. Molecular genetic ap- 
proaches in Drosophila are used to address these 
issues. Ongoing cellular analysis provides increas- 
ing knowledge about the pattern and identity of glia 
and neurons and the projections and pathways taken 
by neuronal growth cones. At the same time, mono- 
clonal antibodies and other molecular probes have 
been generated that have revealed many of these 
cells and their processes (and have also been used to 
identify important genes) . Based on this knowledge 
of the cell biology and availability of specific 
probes, members of the Goodman laboratory have 
been conducting a series of systematic mutant 
screens in order to identify genes that control 
growth cone guidance and target recognition in the 
Drosophila embryo. 
Mutations That Affect Growth Cone 
Guidance in the Developing CNS 
The first large-scale (near-saturation) screen for 
mutations that affect growth cone guidance was un- 
dertaken using a monoclonal antibody that reveals 
all central nervous system (CNS) axon pathways. 
Embryos from over 13,000 mutagenized and bal- 
anced lines were screened, and over 250 mutations 
were saved whose phenorypes suggest possible de- 
fects in growth cone guidance. 
Mutations That Afifect Guidance Toward 
or Away from the Midline 
The growth cones of many CNS neurons initially 
head straight toward the midline, whereas other 
growth cones stay on their own side, leading to the 
hypothesis that midline cells play a key role in the 
formation of the axon commissures. The specificity 
and uniqueness of mutations in two genes, commis- 
sureless (comm) and roundabout (robo), make 
them excellent candidates to encode important com- 
ponents of mechanisms that guide growth cones to- 
ward or away from the midline. In both mutants the 
midline cells are present, and there is no indication 
of cell fate changes in the CNS. Rather, the primary 
defects appear to be in growth cone guidance. 
Mutations in the comm gene lead to a CNS that 
lacks nearly all commissural pathways, even though 
longitudinal and peripheral pathways, sensory neu- 
rons, and muscles appear normal. In comm mutant 
embryos, the growth cones of commissural neurons 
do not project across the midline but instead extend 
only on their own side of the CNS. 
Mutations in the robo gene lead to a specific and 
complementary misrouting, such that some growth 
cones that normally extend only on their own side of 
the CNS now project across the midline in one of the 
commissures. This mutant phenotype suggests that 
robo is part of a mechanism for keeping certain 
growth cones on their own side. There is consider- 
able precedent in other organisms for repulsive sig- 
nals for growth cone guidance. One model to ex- 
plain these mutant phenotypes is that guidance at 
the midline relies on both attractive and repulsive 
cues. The comm gene product might be a compo- 
nent of an attractive signaling system; the robo gene 
product might be a component of a repulsive signal- 
ing system. 
Mutations That Afifect Guidance 
or Fasciculation in Longitudinal Pathways 
Mutations have also been recovered that affect 
longitudinal axon pathways. Mutations in longitu- 
dinals lacking lead to embryos lacking most longi- 
tudinal axon pathways, although both commissural 
and peripheral pathways are present. The growth 
cones that normally pioneer longitudinal pathways 
initially extend but then stall, even though the lon- 
gitudinal glia are present. 
Other mutations display greater specificity for 
subsets of longitudinal pathways. For example, mu- 
tations have been recovered that affect the forma- 
tion of the MPl pathway, either causing the growth 
cone to stall or, in other mutations, to follow the 
wrong pathway. 
Another gene, fasciclin II (fas II), plays a role in 
selective fasciculation of axons in the MPl pathway. 
Initially identified from a monoclonal antibody 
screen, the fasciclin II protein is a member of the 
immunoglobulin superfamily and can function as a 
homophilic cell adhesion molecule. The protein is 
expressed on axons in the MPl pathway and later on 
several other longitudinal axon pathways and on 
most motoneurons. In fas II mutant embryos, the 
pCC, vMP2, MPl, and dMP2 growth cones extend 
along the longitudinal glia but fail to fasciculate and 
do not form a tight bundle of axons. 
The GAL4 enhancer trap method is being used to 
misexpress fasciclin II. When fasciclin II is ectopi- 
cally misexpressed on sensory neurons in the periph- 
400 
