provides a genetic approach to dissection of the mo- 
lecular components of signaling pathways that regu- 
late developmental processes in this system. 
Molecular Basis of Friend Erythroleukemia 
Cancer in both humans and experimental animals 
is a multistage process involving the activation of 
dominant-acting oncogenes and the inactivation of 
tumor-suppressor genes. The complex nature of this 
process raises many questions concerning the inter- 
actions among these genes and their protein prod- 
ucts that lead to unregulated cell growth. The 
erythroleukemias induced by the various strains of 
the murine retrovirus Friend leukemia virus have 
provided an excellent experimental system in 
which to study the multistage nature of cancer. 
By studying the molecular changes in leukemic 
clones that emerge late after viral infection, it be- 
came evident that inactivation of the p53 gene is a 
frequent and probably obligate step in the evolution 
of Friend leukemia. The conclusion that p53 is a 
tumor-suppressor gene has been confirmed and ex- 
tended by recent experiments demonstrating muta- 
tion and allelic loss in a broad spectrum and high 
proportion of human cancers and by the finding that 
wild-type p53 can inhibit the transforming ability of 
mutant p53. 
In addition to inactivation of the p53 gene, activa- 
tion of one of two novel members of the ets gene 
family of DNA-binding transcriptional activators oc- 
curs during the evolution of Friend leukemia. The 
ets gene Spi-1 (spleen focus-forming virus pre- 
ferred integration site 1 ) is activated by proviral in- 
sertion in 95% of leukemic clones induced by 
Friend, whereas another novel ets gene, Fli-1, is ac- 
tivated by insertion of the replication-competent 
Friend murine leukemia virus. The strict specificity 
of integration sites suggests that Fli- 1 and Spi- 1 are 
functionally distinct, encoding proteins that transac- 
tivate a distinct set of genes downstream in the leu- 
kemogenic pathway. Consistent with this possibil- 
ity. Dr. Bernstein and his colleagues have shown that 
DNA-binding ETS domains of Fli- 1 and Spi- 1 recog- 
nize distinct DNA sequences and transactivate dis- 
tinct transcriptional regulatory regions in in vitro 
expression studies. Experiments are under way to 
generate animals harboring mutations in either the 
Fli-1 or Spi-1 genes, to determine their phenotype 
and assess their susceptibility to erythroleukemia 
induction. 
The white-spotting and Steel Loci 
Mutations in mice at either the dominant white- 
spotting (VT) or Steel {St) loci can lead to coat color 
defects, severe macrocytic anemia, and sterility. 
Consistent with the intrinsic and extrinsic nature of 
the Wand SI phenotypes, respectively, \F encodes 
the Kit receptor tyrosine kinase and SI encodes its 
ligand. As expected from earlier biological analysis 
of VFand SI mutants, c-kit is expressed in early hema- 
topoietic progenitor/stem cells — melanoblasts, me- 
lanocytes, and primordial germ cells. SI is expressed 
in cells that immediately surround those expressing 
c-kit, including fibroblasts, Sertoli cells, and granu- 
losa cells. Thus the Kit signaling pathway appears to 
be activated by cell-cell contact, a conclusion sup- 
ported by the observation that the Sl'^ mouse mu- 
tant, which fails to make the membrane-bound form 
of the Steel protein, has a mutant SI phenotype. 
This close apposition of cells expressing T^^and SI 
is found not only in the three cell lineages affected 
by Wor SI mutations but also, for example, in the 
brain, where Q-kit and SI are found at very high lev- 
els in the hippocampus and cerebellum. These ob- 
servations suggest that IFand SI might play a critical 
role in neural function — in neural development, 
axonal guidance, and/or higher neural functions. 
Mice homozygous for severe SI or IFpoint mutations 
display no obvious abnormality or deficit of c-kit- 
expressing neuronal cells in the brain, suggesting 
that this signaling pathway plays a postmitotic role 
in the central nervous system. 
Together these approaches to cell signaling 
should contribute to emerging concepts concerning 
the molecular basis of cell-cell communication dur- 
ing normal embryological development, hema- 
topoiesis, neural function, and malignancy. 
Dr. Bernstein is Associate Director and Head of 
the Division of Molecular and Developmental Biol- 
ogy of the Samuel Lunenfeld Research Institute of 
Mount Sinai Hospital, Toronto, and Professor of 
Molecular and Medical Genetics at the University 
of Toronto. 
Books and Chapters of Books 
Breitman, M.L., and Bernstein, A. 1992. Engineer- 
ing cellular deficits in transgenic mice by genetic 
ablation. In Transgenic Animals (Grosveld, F., 
and Kallias, G., Eds.). San Diego, CA: Academic, 
pp 127-146. 
Reith, A., and Bernstein, A. 1991 . Molecular biol- 
ogy of the W^and Steel loci. In Genome Analysis: 
Genes and Phenotypes (Davies, K.E., and Tilgh- 
man, S.M., Eds.). Cold Spring Harbor, NY: Cold 
Spring Harbor, vol III, pp 105-133- 
INTERNATIONAL RESEARCH SCHOLARS 501 
