PROTEIN STRUCTURE AND FUNCTION 
Florante a. Quiocho, Ph.D., Investigator 
The three-dimensional structures of a number of 
proteins are being determined by x-ray crystallogra- 
phy in order to gain detailed molecular under- 
standing of the functions of these proteins— espe- 
cially the role of molecular recognition— and the 
biological processes associated with these func- 
tions. 
I. Adenosine Deaminase. 
Adenosine deaminase is of particular interest be- 
cause impairment of its catalytic function is associ- 
ated with severe combined immunodeficiency, af- 
fecting both T and B lymphocytes. Last year Dr. 
Quiocho reported the crystallization of two com- 
plexes of the enzyme— one with 2'-deoxycofor- 
mycin, a potent inhibitor and potential antilympho- 
cytic compound, and the other with purine 
riboside, a ground-state analogue. In the past year, 
heavy-atom searching for phase determination of 
the crystal of the purine riboside complex has 
yielded four derivatives. Electron density maps have 
o 
been calculated at 3.2 A resolution, using phases 
obtained by multiple isomorphous replacement 
and by iterative single isomorphous replacement, 
which takes advantage of the crystal's high solvent 
content. These maps, which are similar, clearly re- 
vealed the intermolecular boundary. Interpretation 
of these maps is under way. 
II. Antibody. 
Dr. Quiocho and his colleagues have obtained 
crystals of Fab of a BAT123 murine monoclonal an- 
tibody raised against the gpl20 coat protein of the 
HTLV-IIIB strain of human immunodeficiency virus. 
BAT123 binds to gpl20 with high affinity {K ~ 1.4 
10—1 ^ 
X 10 M ) at a site considered to be the most sen- 
sitive for antibody neutralization of the virus. The 
antibody was provided by Dr. T. W Chang (Baylor 
College of Medicine). 
III. Periplasmic Binding Proteins. 
Binding proteins, which serve as initial receptors 
for bacterial active transport and chemotaxis, are 
ideal for detailed structure and function studies. 1) 
They can be easily purified in large quantities (0.2 
to 2 g). 2) They are extremely stable and can be 
crystallized easily in excellently diffracting forms. 
3) They bind diverse sets of ligands, such as carbo- 
hydrates, amino acids, and oxyanions. 4) Atomic 
structures of several binding proteins, in the 
liganded or unliganded forms, have been deter- 
mined. 5) Several different binding protein genes 
have been cloned. The following studies have been 
initiated recently. 
A. Site-directed mutagenesis. Site-directed muta- 
genesis provides a powerful approach to dissecting 
the roles of the various regions of these proteins, 
down to specific amino acid side chains, especially 
in structural integrity and biological function. With 
the availability of the highly refined, extremely high 
resolution structures of the arabinose-binding pro- 
tein complexed with different sugar substrates and 
with the cloning of the gene coding for the protein 
in an Escherichia coli overproducer strain, the 
sugar-binding site region is especially suited for mu- 
tagenesis studies. This region is composed of four 
types of residues: type I, those that directly hydro- 
gen bond sugar substrates; type II, those that indi- 
rectly, via bound water molecules, form hydrogen 
bonds with the substrates; type III, those that have 
nonpolar interactions with the ligand; and type ly 
those that are located in the hinge between the two 
domains and form the base of the sugar-binding site 
cleft between the two domains. Dr. Vermersch has 
obtained the following results. 1) Substitution of 
Asp90, a type I residue, by Glu almost completely 
abolishes sugar-binding activity. 2) Replacement of 
type II residues, such as Glnll by Asn and Thrl47 
by Gly only diminishes sugar binding by about a 
half 3) Considerable enhancement in ligand bind- 
ing was achieved by substituting MetlOS, a type III 
residue, with Leu. 4) Substitution of Pro254, a type 
IV residue, by Gly also resulted in an enhancement 
of binding activity Excellently diffracting crystals of 
all these mutant proteins have been obtained, pav- 
ing the way for molecular understanding of the ef- 
fects of these mutations. 
B. Substrate engineering. The combination of crys- 
tallographic and binding studies for a number of 
deoxygenated and fluoro-substituted analogues of 
sugar substrates of the arabinose-binding protein is 
providing new insights into protein-ligand interac- 
tions. For example, D-galactose is only about two- 
fold less tightly bound than L-arabinose, and all 
polar groups of both sugars interact with the pro- 
Continued 
593 
