Bacteria expressing R67-plasmid encoded dihydrofolate reductase (R67 DHFR) exhibit high-level resistance to the antibiotic trimethoprim. Native R67 DHFR is a 34,000 M(r) homotetramer which exists in equilibrium with an inactive dimeric form. The structure of native R67 DHFR has now been solved at 1.7 A resolution and is unrelated to that of chromosomal DHFR. Homotetrameric R67 DHFR has an unusual pore, 25 A in length, passing through the middle of the molecule. Two folate molecules bind asymmetrically within the pore indicating that the enzyme's active site consists of symmetry related binding surfaces from all four identical units.

The binary complex of rC and adenosine adopts an 'intermediate' conformation relative to the previously described 'closed' and 'open' conformations of other rC complexes. Based on a comparison of these structures, the induced fit that is necessary for catalysis and closing of the active-site cleft appears to be confined to the small lobe, as in the absence of the peptide the conformation of the large lobe, including the peptide-docking surface, does not change. Three specific components contribute to the closing of the cleft: rotation of the small lobe; movement of the C-terminal tail; and closing of the so-called glycine-rich loop. There is no induced fit in the large lobe to accommodate the peptide and the closing of the cleft. A portion of the C-terminal tail, residues 315-334, serves as a gate for the entry or exit of the nucleotide into the hydrophobic active-site cleft.

The crystal structure of the hexahistidine-tagged mouse recombinant catalytic subunit (H 6 -rC) of cAMP-dependent protein kinase (cAPK), complexed with a 20-residue peptide inhibitor from the heatstable protein kinase inhibitor PKI(5-24) and adenosine, was determined at 2.2 Å resolution. Novel crystallization conditions were required to grow the ternary complex crystals. The structure was refined to a final crystallographic R-factor of 18.2% with good stereochemical parameters. The "active" enzyme adopts a "closed" conformation as found in rC:PKI(5-24) [Knighton et al. (1991a,b) Science 253, 407-414, 414-420] and packs in a similar manner with the peptide providing a major contact surface. This structure clearly defines the subsites of the unique nucleotide binding site found in the protein kinase family. The adenosine occupies a mostly hydrophobic pocket at the base of the cleft between the two lobes and is completely buried. The missing triphosphate moiety of ATP is filled with a water molecule (Wtr 415) which replaces the γ-phosphate of ATP. The glycine-rich loop between 1 and 2 helps to anchor the phosphates while the ribose ring is buried beneath -strand 2. Another ordered water molecule (Wtr 375) is pentacoordinated with polar atoms from adenosine, Leu 49 in -strand 1, Glu 127 in the linker strand between the two lobes, Tyr 330, and a third water molecule, Wtr 359. The conserved nucleotide fold can be defined as a lid comprised of -strand 1, the glycine-rich loop, and -strand 2. The adenine ring is buried beneath -strand 1 and the linker strand (120-127) that joins the small and large lobes. The C-terminal tail containing Tyr 330, a segment that lies outside the conserved core, covers this fold and anchors it in a closed conformation. The main-chain atoms of the flexible glycinerich loop (residues 50-55) in the ATP binding domain have a mean B-factor of 41.4 Å 2 . This loop is quite mobile, in striking contrast to the other conserved loops that converge at the active site cleft. The catalytic loop (residues 166-171) and the Mg 2+ positioning loop (residues 184-186) are a stable part of the large lobe and have low B-factors in all structures solved to date. The stability of the glycine-rich loop is highly dependent on the ligands that occupy the active site cleft with maximum stability achieved in the ternary complex containing Mg‚ATP and the peptide inhibitor. In this ternary complex the γ-phosphate is secured between both lobes by hydrogen bonds to the backbone amide of Ser 53 in the glycine-rich loop and the amino group of Lys 168 in the catalytic loop. In the adenosine ternary complex the water molecule replacing the γ-phosphate hydrogen bonds between Lys 168 and Asp 166 and makes no contact with the small lobe. This glycine-rich loop is thus the most mobile component of the active site cleft, with the tip of the loop being highly sensitive to what occupies the γ-subsite.

Protein kinases are essential for the regulation of cellular growth and metabolism. Since their dysfunction leads to debilitating diseases, they represent key targets for pharmaceutical research. The rational design of kinase inhibitors requires an understanding of the determinants of ligand binding to these proteins. In the present study, a theoretical model based on continuum electrostatics and a surface-area-dependent nonpolar term is used to calculate binding affinities of balanol derivatives, H-series inhibitors, and ATP analogues toward the catalytic subunit of cAMP-dependent protein kinase (cAPK or protein kinase A). The calculations reproduce most of the experimental trends and provide insight into the driving forces responsible for binding. Nonpolar interactions are found to govern protein-ligand affinity. Hydrogen bonds represent a negligible contribution, because hydrogen bond formation in the complex requires the desolvation of the interacting partners. However, the binding affinity is decreased if hydrogen-bonding groups of the ligand remain unsatisfied in the complex. The disposition of hydrogen-bonding groups in the ligand is therefore crucial for binding specificity. These observations should be valuable guides in the design of potent and specific kinase inhibitors.