domain (which is characteristic of the DNA-binding 
domains from a large family of receptors) contains 
two 4S zinc fingers. Cocrystallization experiments 
are well under way with complexes containing the 
DNA target. 
Dr. Steven McKnight (HHMI, The Carnegie Insti- 
tution of Washington) and subsequently others 
»-have identified a family of transcription factors that 
dimerize through a domain called a leucine zipper 
(leucine heptad repeat), which has been shown to 
be a parallel coiled coil. Drs. Norma Duke and Jon 
Shuman have crystallized a complex formed by a 
homodimeric 20 kDa construct and its DNA target. 
Recently Dr. Sigler and his colleagues purified a 
62-amino acid construct derived by Tao Pan and 
Dr. Joseph Coleman (Yale University) from the 
GAL4 DNA-binding domain and expressed from an 
overproducing Escherichia coli clone. Like the 
glucocorticoid receptor, this DNA-binding domain 
contains a 4S zinc finger. 
The DNA-binding domain of the E2 transactivator 
of bovine papilloma virus has been overexpressed 
from an E. coli clone (engineered by Steve Gross- 
man of The University of Chicago) and purified. Ef- 
forts are under way to crystallize the DNA-binding 
domain both alone and in cocrystalline complexes 
with the E2 enhancer sequence. 
The TATA sequence and its binding factor TFIID 
define the position and polarity of the vast majority 
of PolII initiation complexes. Dr. Steven Hahn (orig- 
inally at the Massachusetts Institute of Technology), 
in collaboration with Dr. Phillip Sharp, purified and 
cloned TFIID from yeast. Efforts are under way to 
crystallize the purified protein and its TATA com- 
plex from material that Dr. Hahn has recently puri- 
fied from overexpressing E. coli constructs. 
II. Translational Control. 
The crystal structure of yeast initiator tRNA has 
o 
been refined to 3 0 A with imaging-plate data from 
the Photon Factory. Several functionally important 
features emerged: A20, A59, and A60 form part of a 
unique and consistent sequence feature of eu- 
karyotic initiator tRNA. They combine to make a 
unique hydrogen-bonded arrangement that in- 
volves the shortened backbone of the seven- 
membered D loop, another characteristic of this 
functional class. Dr. Thomas Wagner has shown 
that this contiguous region is the stereochemical 
reason why eukaryotes reject Met-tRNA.'^" in the 
peptide elongation process. 
Genetic studies (Dr. U. RajBhandary, Massachu- 
setts Institute of Technology) confirmed that the 
features most responsible for initiator activity are in 
the anticodon arm. The structure of the anticodon 
loop in yeast initiator tRNA is different from that of 
crystalline elongator tRNAs; however, the antico- 
don loop is involved in a crystal packing contact, 
and therefore one cannot be sure whether this con- 
formational difference is intrinsic to the structure 
or imposed by the contact. Moreover, the antico- 
don arm is the most poorly ordered element in the 
structure, thereby obscuring the stereochemical 
features of the three consecutive G-C^ base pairs at 
the distal end of the stem. Crystallographic and nu- 
clear magnetic resonance (NMR) studies will be car- 
ried out on other initiator tRNAs and synthetic RNA 
fragments to clarify these issues further. 
III. Transmembrane Signaling. 
Receptor-mediated signal transduction involves 
the release of second messengers, either through 
the synthesis of cyclic nucleotides or the phos- 
pholipase-catalyzed liberation of hydrolysis prod- 
ucts from phospholipids. These include arachidon- 
ic acid, the precursor of the eicosanoid mediators 
of inflammation. Inositol trisphosphate and 
diacylglycerol are released by phospholipase C ac- 
tion on phosphoinositides. 
The chemistry of phospholipases is interesting, 
since the natural substrate is a lamellar or micellar 
aggregate of phospholipids rather than a soluble 
dispersed individual molecule. The mechanism of 
enzymatic attack on an aggregated substrate at the 
solvent-bilayer interface presents a special chal- 
lenge to understanding transmembrane signaling in 
molecular terms. 
A. Crystallographic studies. Dr. Sigler and his col- 
leagues have now solved and refined a series of 
crotalid venom phospholipase A^ (PLA^) structures 
to 2.0 A resolution or better. Many of the biochemi- 
cal barriers to formulating a mechanism may have 
been leapfrogged by the crystallization of at least 
one and possibly more PLA^s in complexes with a 
phosphonate transition-state analogue. Since one 
such complex is in the early stages of crystallo- 
graphic refinement against 1.9 A data, it is likely 
that the stereochemical mechanism for catalysis will 
be seen at the membrane-liquid interface. 
B. Site-directed mutagenesis. Starting ft-om syn- 
thetic sticky-ended oligonucleotide duplexes of 
20-30 base pairs, Sandra Luisi has synthesized 
Continued 
602 
