screening procedures use the polymerase chain re- 
action to amplify any binding sites that occur in a 
pool of random synthetic fragments.) Experiments 
with these peptides may help determine the precise 
DNA-binding site for a zinc finger domain and 
should provide suitable molecules for cocrystalliza- 
tion. The tandem peptides should also be a useful 
starting point for the design of zinc finger proteins 
with novel specificities. 
III. Studies of the tat and rev Proteins from HIV 
The key regulatory proteins from Hiy which is 
the etiological agent of AIDS, are tat and rev. Al- 
though the mechanisms of action are not fully un- 
derstood, tat appears to control transcription 
and/or translation, while rev controls splicing 
and/or nuclear export of the messenger RNA. Both 
proteins are required for viral replication, and 
these are attractive targets for drug design. 
The laboratory has pursued several structure and 
activity studies of the tat protein. In the most re- 
cent work, Dr. Alan Frankel showed that tat could 
be taken up by cells in tissue culture, be trans- 
ported to the nucleus, and activate expression of a 
reporter gene. Although the biological significance 
of this is unclear, the study raised important ques- 
tions about the mechanism of tat action and sug- 
gested new approaches for biochemical studies. 
Detailed structural analysis might give clues for 
drug design, but the laboratory has been unable to 
crystallize tat, and NMR studies done in collabora- 
tion with Dr. Gerhard Wagner (University of Michi- 
gan) indicate that most of purified tat protein exists 
as a random coil in solution. Crystallization at- 
tempts are now focusing on the rev protein (pro- 
PUBLICATIONS 
vided by Dr. Craig Rosen, Roche Institute for Molec- 
ular Biology). 
IV Computer Software for Analysis and Design of 
DNA-binding Proteins. 
The laboratory is continuing to develop a 
database of protein-DNA interactions. The database 
includes site-specific DNA-binding proteins for 
which both the sequence of the protein and the se- 
quence of the binding site are known. As more in- 
formation becomes available about protein-DNA 
recognition, this database should be extremely use- 
ful for model-building studies and for protein de- 
sign. 
The laboratory also is developing strategies for 
computer-aided protein design. These strategies are 
implemented in PDB-PROTEUS, a package of pro- 
grams and subroutines for computer-aided protein 
design. The programs allow one to test a large 
number of sequences or conformations when plan- 
ning sequence changes or designing a new protein. 
Explicit criteria are used for picking the best ar- 
rangements, and these are evaluated by using de- 
tailed energy calculations and examining them on a 
computer graphics system. These programming 
strategies have recently been used to design tight- 
binding variants of the X repressor. Several of these 
variants are being prepared for experimental tests, 
and this work should provide the basis for system- 
atically designing repressors with altered specifici- 
ties. 
Dr. Pabo is also Associate Professor of Molecular 
Biology and Genetics and of Biophysics at The 
Johns Hopkins University School of Medicine. 
Articles 
Frankel, A.D., Chen, L., Cotter, R.J., and Pabo, CO. 1988. Dimerization of the tat protein from human immu- 
nodeficiency virus: a cysteine-rich peptide mimics the normal metal-linked dimer interface. Proc Natl Acad 
Set USA 85:6297-6300. 
Frankel, A.D., and Pabo, CO. 1988. Cellular uptake of the tat protein from human immunodeficiency virus. 
Ce// 55: 1189-1 193^ 
Jordan, S.R., and Pabo, CO. 1988. Structure of the lambda complex at 2.5 A resolution: details of the repres- 
sor-operator interactions. Science 242:893-899. 
Stearman, R.S., Frankel, A.D., Freire, E., Liu, B. , and Pabo, CO. 1988. Combining thermostable mutations in- 
creases the stability of X. repressor. Biochemistry 27:7571-7574. 
592 
