between loss of a specific gene with specific pheno- 
typic consequences, and this thereby establishes the 
function of that gene in development. By comparing 
the phenotypic consequences of mutating a series of 
genes in a developmental network, it is hoped that a 
logic of how that network functions will emerge. 
The laboratory of Investigator Shirley M. Tilgh- 
man, Ph.D. (Princeton University) is also studying 
the genes responsible for appropriate development 
of the mouse. This year her laboratory proposed a 
model to explain how two neighboring genes could 
influence each other's expression by competing for 
a common set of regulatory elements. The conse- 
quence of the competition is that only one of the 
two genes is expressed from a single chromosome, a 
phenomenon known as parental imprinting. Her lab- 
oratory also demonstrated that the repression of the 
a-fetoprotein gene in liver is a complex process, in 
that the regulatory elements necessary for repres- 
sion in hepatocytes around the central vein are dif- 
ferent from those in the intermediate and periportal 
hepatocytes. The laboratory is using classical genet- 
ics to map two genes in which mutations cause de- 
fects in development. Mutations in the Fused gene 
result in an overgrowth of neuroectoderm, whereas 
piebald mutations affect the development of two 
neural crest lineages, the melanocytes and enteric 
ganglia. The genetic analyses will be important in 
strategies to clone these genes. 
Research in the laboratory of Assistant Investiga- 
tor Philippe M. Soriano, Ph.D. (Baylor College of 
Medicine) continues to concern developmental ge- 
netics in mice. They have produced mutant mice for 
several nonreceptor tyrosine kinases, using gene- 
targeting techniques in embryonic stem cells. These 
kinases play essential roles in transducing signals 
from receptors with extracellular domains. Whereas 
disruption of the src gene in mice leads to the bone 
disease osteopetrosis, disruption of other related ki- 
nase genes does not lead to overt phenotypes. Stud- 
ies of the fyn~ mice have revealed defects in 
signaling from the T cell receptor and in the hippo- 
campus, leading to defects in the immune system 
and in learning and memory, respectively. Crosses 
between kinase mutant mice lead to novel pheno- 
types, embryonic death or glomerulonephritis in 
the case of animals deficient for both yes and fyn. 
Another aspect of Dr. Soriano's work involves the 
generation of mouse mutations using "promoter 
traps," in which the gene is both mutated and its 
activity traced by expression of an introduced pro- 
moterless reporter gene. These studies have led to 
the identification of 18 new recessive lethal strains, 
of which one is a mutation in a broadly expressed 
transcription factor, TEFl. 
The laboratory of Investigator Richard D. Pal- 
miter, Ph.D. (University of Washington), in collabo- 
ration with Dr. Ralph Brinster at the University of 
Pennsylvania, uses gene transfer into the germline of 
mice to study various aspects of development and 
disease. The genes of interest are injected into pro- 
nuclei of fertilized eggs, where they integrate into 
one of the mouse chromosomes and become new 
genetic traits. By combining the regulatory regions 
(promoters and enhancers) from one gene with the 
region that encodes a particular protein from an- 
other, it is possible to direct the expression of novel 
gene products to specific cell types. This approach 
is currently being used to study the development of 
the mammalian nervous system. For example, the 
expression of nerve growth factor was directed to 
sympathetic neurons using the promoter/enhancer 
from a gene required for synthesis of their neuro- 
transmitter, norepinephrine. This showed that 
growth factor gradients are not required to guide 
sympathetic axons to their targets, but are required 
to establish the normal pattern of innervation within 
the target tissue. The small population of neurons 
that normally use epinephrine as their neurotrans- 
mitter have been eliminated from a line of mice, by 
expressing diphtheria toxin using the promoter/ 
enhancer from a gene involved in epinephrine syn- 
thesis. Studies of these mice now provide a means 
for ascertaining normal neuron function. Methods 
are also being devised to alter the neurotransmitters 
that neurons normally make to explore further the 
effects of these molecules. 
The laboratory of Assistant Investigator Gregory S. 
Barsh, M.D., Ph.D. (Stanford University) is studying 
the mouse agouti {A) locus. Some dominant A mu- 
tations (y4^, A"^) produce obesity and liver tumors in 
the animals, and some recessive mutations {a") are 
lethal during early embryogenesis. A candidate gene 
for A, ASP, has been isolated that appears to encode 
a small secreted protein, and a cell culture system 
has been developed to study increased tumor sus- 
ceptibility mediated by A''^. Molecular alterations 
have been identified in A^ and in a^ that appear to 
affect the expression but not the structure of ASP. 
Studies are currently under way to deterrriine if and 
how different agouti mutations affect pigment syn- 
thesis, obesity, tumor formation, and early devel- 
opment. 
The goal of the laboratory of Assistant Investigator 
Jeffrey M. Friedman, M.D., Ph.D. (Rockefeller Uni- 
versity) is to clone genes that, when defective, give 
rise to obesity. Extensive studies of human twins 
have suggested that body weight in a given individ- 
ual is physiologically regulated and that each per- 
son's weight is determined in part by a set of specific 
GENETICS 143 
