MOLECULAR STUDIES OF BEHAVIOR AND DEVELOPMENT IN DROSOPHILA 
Michael W. Young, Ph.D., Investigator 
I. Neurogenesis. 
A study of the genetic control of ectoderm devel- 
opment in Drosophila has been initiated through 
the recovery of neurogenic mutants. These over- 
produce neuroblasts at the expense of nonneural 
ectoderm. The best-characterized mutants in this 
series are associated with complete loss of function 
at the Notch locus. From its DNA sequence, Notch 
appears to encode a 288 kDa protein, the structure 
of which is dominated by a 36-fold tandem repeti- 
tion of a cysteine-rich peptide related to epidermal 
growth factor (EGF). In Notch, no two of these re- 
peats are identical, and genetic and molecular anal- 
yses of mutations affecting this domain of the pro- 
tein indicate that differentiation of function exists 
among the different EGF-like segments. Genetic in- 
teractions between Notch alleles and between 
Notch and unlinked neurogenic loci can be modi- 
fied by amino acid substitutions affecting specific 
EGF-like repeats of the Notch protein. All of this is 
consistent with the notion that Notch participates 
in ectodermal development through cell-cell inter- 
actions. 
Dr. Young and his colleagues recently character- 
ized the Notch protein itself biochemically and de- 
termined its distribution in developing Z)ro5qp^//«. 
Antibodies show that Notch is a stable, high-molec- 
ular-weight glycoprotein. This transmembrane pro- 
tein, with the EGF-like elements exposed at the cell 
surface, is phosphorylated variably on serines of the 
cytoplasmic domain. Individual Notch polypeptide 
chains appear to be joined to each other by disul- 
fide bonds, suggesting that homotypic interaction 
of these proteins is required for function. It is not 
known whether these interactions generally involve 
molecules on the same or neighboring cell surfaces. 
Immunocytochemistry has shown that at the 
time of gastrulation. Notch is largely restricted to 
ectodermal cells defining the neurogenic region of 
the embryo. All of these cells have the potential to 
form neuroblasts, but only a fraction do so; the re- 
mainder become precursors of the epidermis in a 
lineage-independent fashion. Thus Notch appears 
to be associated with both epidermal and neural 
precursors. High levels of Notch expression eventu- 
ally become restricted to neuroblasts, but only after 
these begin to delaminate from the embryonic ec- 
toderm and come into close association with meso- 
derm. Mesodermal cells also produce Notch as 
neuroblast delamination proceeds, indicating a 
possible role in embryonic neurogenesis. During 
larval and pupal development. Notch becomes 
preferentially associated with nervous tissue and is 
expressed on stem cells, fully differentiated cells, 
and fasciculating nerve processes. Thus Notch 
seems to be present before and long after cell fate 
is determined. In the eye imaginal disk. Notch is 
switched on in the morphogenetic furrow. This is 
where cells that will compose ommatidia aggregate 
prior to establishing their specific developmental 
fates. As in embryonic neurogenesis, fate is deter- 
mined by position, not lineage, and all cells make 
Notch. The protein continues to be associated with 
ommatidia after differentiation. Uniform expression 
on cells interacting to produce different develop- 
mental lineages from single primordia in the eye 
and embryonic ectoderm suggests that Notch alone 
may be insufficient to elaborate cell fates. 
II. Biological Rhythms. 
Periodic functions from heartbeat to 24 h (circa- 
dian) sleep/wake cycles are affected by mutations 
at the per locus. The 1,200-amino acid protein 
encoded by per shows some sequence similarity 
to vertebrate proteoglycans. Changes in a fly's 
rhythms can be produced by amino acid substitu- 
tions in the per protein, or by regulating per ex- 
pression. In transgenic Drosophila a relationship 
between period length and abundance of the gene 
products is seen: high levels of expression lead to a 
short-period biological clock; low levels are linked 
to long-period behavior. 
Cell-level phenotypes have been recognized for 
per mutants in the laboratory through studies of in- 
tercellular junctional communication in salivary 
glands. Low levels of per expression, and arrhyth- 
mic and long-period phenotypes, are . correlated 
with poor intercellular junctional communication. 
Alternately, higher than wild-type levels of cell- 
to-cell communication are observed in mutant tis- 
sue from Drosophila having fast running biological 
clocks. 
Cells making per protein in developing Drosoph- 
ila have been identified by in situ hybridization to 
per transcripts and immunocytochemistry. Very low 
levels of protein are found in segmentally arranged 
clusters of cells composing part of the embryonic 
nervous system. Later in development, per RNA and 
Continued 
343 
