BODY PATTERNING IN DROSOPHIIA EMBRYOS 
Gary Struhl, Ph.D., Assistant Investigator 
Dr. Struhl's research is focused on the molecular 
nature and mode of action of spatial determinants 
responsible for segregating the early Drosophila 
embryo into its characteristic pattern of body seg- 
ments. Both classic embryological and genetic ex- 
periments have shown that the segment pattern is 
prefigured at fertilization by distinct anterior, poste- 
rior, and terminal determinant systems laid down in 
the egg during oogenesis. The roles of each of 
these systems, as well as the molecular mechanisms 
involved in their initial establishment and subse- 
quent function, have been investigated. 
I. Anterior Determinant System. 
Prior work by Dr. Christiane Niisslein-Volhard 
and her colleagues has established that a single 
morphogen, bicoid {bed), controls anterior body 
pattern. Transcripts of the gene are synthesized in 
the nurse cells and then transported to the oocyte, 
where they appear to be trapped at their point of 
entry— the future anterior end of the embryo. After 
fertilization these tightly localized transcripts are 
translated, and the resulting protein diffuses from 
its site of synthesis, generating a concentration gra- 
dient. The bed gradient then dictates anterior body 
pattern, by controlling where subordinate regula- 
tory molecules are expressed. 
Recent studies in this laboratory have shown that 
the 3 '-noncoding portion of the bed transcript con- 
tains a cis-acting signal that is both necessary and 
sufficient to allow mRNAs synthesized in the nurse 
cells to be trapped selectively upon entry at the an- 
terior pole of the oocyte. This large (625 bp) local- 
ization signal is likely to form an extensive second- 
ary structure, suggesting that it acts as a ligand that 
is bound and anchored by a receptor protein in the 
oocyte. Thus establishment of the bed morphogen 
gradient may depend on the selective retention of 
bed transcripts by specialized receptor proteins as 
the transcripts pass from the nurse cells into the 
oocyte. 
The bed protein contains a homeobox domain, a 
structural motif associated with site-specific DNA 
binding in prokaryotes, yeast, and higher eukary- 
otes, suggesting that it might function as a tran- 
scription factor. Hence the graded distribution of 
bed protein might determine where subordinate 
regulatory genes are expressed, by virtue of its abil- 
ity to activate or repress transcription directly in a 
concentration-dependent fashion. Dr. Herbert 
Jackie and his colleagues have recently shown that 
the regulatory gene hunehbaek (hb) is normally ac- 
tivated in a 6c<i-dependent fashion in a broad but 
sharply bounded anterior domain in early embryos, 
suggesting that it may respond to the bed gradient 
by this mechanism. 
The possibility that the bed morphogen acts as a 
concentration-dependent transcriptional regulator 
has been tested directly by creating a series of hy- 
brid genes in which small portions of the upstream 
regulatory region of hb have been inserted in front 
of a naive transcriptional start site, driving expres- 
sion of the coding sequence for P-galactosidase. 
When returned to the genome, a subset of these 
hybrid genes were expressed in sharply bounded 
anterior domains specified by the bed gradient (like 
the endogeneous hb gene). Ml of the hybrid genes 
in this subset contained specific cis-acting elements 
found within a small portion (250 bp) of the hb up- 
stream regulatory region. Furthermore, it was possi- 
ble to reconstruct bed-dependent transcriptional 
activation mediated by these same elements in 
yeast, a heterologous system lacking all other Dro- 
sophila components. These studies provide a 
strong argument that direct interactions between 
bed protein and discrete cis-acting regulatory ele- 
ments in the hb gene are responsible for control- 
ling where the hb gene is normally expressed dur- 
ing early embryogenesis. It was also shown that the 
posterior boundaries of expression of the various 
hybrid genes were determined by the number and 
quality of the cis-acting elements mediating their re- 
sponse to bed protein. This result provides direct 
evidence that the bed gradient can dictate several 
distinct domains of anterior gene expression, de- 
pending on the number or quality of the regulatory 
sites governing the response of particular target 
genes. Preliminary genetic and molecular studies 
have suggested several other candidate genes that 
normally may respond directly to the bed gradient. 
II. Posterior Determinant System. 
Control of posterior body pattern was initially 
thought to occur in a symmetrical fashion to that of 
anterior body pattern, depending on an opposing 
gradient of the morphogen nanos. However, in 
work begun during the previous year, a surprising 
result was obtained that suggested that this view is 
Continued 
545 
