promoter (HS-if^ flies). Ubiquitous wingless expres- 
sion results in naked ventral cuticle, broadening of 
the engrailed domain in the posterior direction, and 
induction of endogenous wingless transcription. 
Jasprien Noordermeer and John Klingensmith have 
combined the HS-wg allele with loss-of-function mu- 
tations in various other segment polarity genes. The 
results of these double-mutant experiments point to 
essential functions for several genes in the wingless 
pathway. For example, it appears that the expression 
of dishevelled is essential to generate the naked cuti- 
cle and the ectopic expression of engrailed. In con- 
trast, armadillo is not required for the broad en- 
grailed expression induced by W?)-wg but is 
necessary to maintain this pattern. John Klingensmith 
and Frank van Leeuwen are currently using specific 
antibodies and transfected cell lines to analyze the 
dishevelled gene product. 
wingless is also part of a gene family in Drosoph- 
ila. One of the wingless-xe\2Xe<\ genes, DWnt-3, is 
highly unusual. Its predicted protein is much longer 
than other Wnt proteins and contains long inserts at 
various sites. Dr. Fradkin has made an antibody spe- 
cific for this member of the Wnt gene family. With 
in situ hybridization methods and with antibody 
stainings, it was found that DWnt-3 is expressed in 
the developing CNS of Drosophila embryos, with 
additional sites in the epidermis of gnathal seg- 
ments. Experiments are under way to assess the role 
of these genes in embryogenesis by isolating mu- 
tants and by overexpression of the genes under the 
control of a heat-shock promoter. 
Dr. Nusse is also Associate Professor of Develop- 
mental Biology at Stanford University School of 
Medicine. 
Articles 
Nusse, R. 1991. Insertional mutagenesis in mouse 
mammary tumorigenesis. Curr Top Microbiol 
Immunol 171:43-65. 
Nusse, R., and Varmus, H.E. 1992. genes. Cell 
69:1073-1087. 
Roelink, H., and Nusse, R. 1992. Using mRNA in 
situ hybridization to localize Wnt-3 and Wnt-3A 
expression in the developing neural tube. Meth- 
ods Neurosci 9:256-273. 
Roelink, H., Wagenaar, E., and Nusse, R. 1992. 
Amplification and proviral activation of several 
Wnt genes during progression and clonal varia- 
tion of mouse mammary tumors. Oncogene 
7:487-492. 
Russell, J., Gennissen, A., and Nusse, R. 1992. Iso- 
lation and expression of two novel Wnt/wingless 
gene homologues in Drosophila. Development 
115:475-485. 
MECHANISM OF DNA REPLICATION 
Michael E. O'Donivell, Ph.D., Assistant Investigator 
Duplication of the genetic material is central to 
the life process of every cell. Although this need 
occur only once, it must be performed accurately to 
preserve the species. Dr. O'Donnell and his col- 
leagues are studying the series of individual steps by 
which the genetic material is duplicated. The sys- 
tem being studied is the bacterium, Escherichia 
coli, which, like most organisms, stores its genetic 
information in the form of double-strand DNA. Du- 
plication of the DNA, a process called replication, 
requires more than a dozen proteins. Ten of these 
are bound together into a multiprotein machine: 
one subunit of this machine is the actual DNA poly- 
merase; another subunit is a 3'- 5' exonuclease that 
proofreads the product of the DNA polymerase. The 
multiprotein chromosomal replicase is called DNA 
polymerase III holoenzyme, referred to here as the 
"holoenzyme." 
Both polymerase and exonuclease activities are 
found in most DNA polymerases, including the 
classic DNA polymerase I of E. coli, which mainly 
functions to repair damage to the DNA chromo- 
some incurred during everyday life. It is the other 
eight subunits, or accessory proteins, of the ho- 
loenzyme that are unique to the chromosomal re- 
plicative polymerase. It seems likely that they 
each have an individual function in the process of 
duplicating a chromosome in which the two 
strands of double-stranded DNA are separated, and 
then each single strand is used as a template to 
make two new daughter duplexes. The process 
sounds simple, but is difficult; too little is known 
even to guess at individual functions for these 
eight accessory proteins. A combination of genet- 
ics and biochemistry is being used to identify the 
steps in the replicative process. 
98 
