Search

The crystallization of leucine aminopeptidase from hog kidney is reported for the first time. The crystals which diffract to 4-A resolution have the space group P2(1)2(1)2(1) (a = 186.3 A, b = 223.2 A, and c = 80.5 A) and contain four hexamers per unit cell, or one per asymmetric unit. Electron micrographic images of hog kidney leucine aminopeptidase are indistinguishable from micrographs of beef leucine aminopeptidase taken under the same conditions (10). These reveal an equilateral triangle of about 85 A per side, seemingly made of three 40-A diameter spheres. This triangle is circumscribed by another concentric, less-dense triangle of 120 A per side which is rotated 60 degrees with respect to the inner triangle. Immunodiffusion and microcomplement fixation assays indicate that the two enzymes share greater than 90% amino acid sequence homology. This similarity is corroborated by peptide maps of tryptic fragments of the radioiodinated enzymes. The model of the quaternary structure proposed to explain the appearance of electron micrographs of single molecule and crystalline bovine lens enzyme also describes the hog kidney enzyme equally well. That the model of leucine aminopeptidase originally proposed for the beef enzyme also can be used to describe hog kidney leucine aminopeptidase crystal packing in the highly anisometric unit cell provides further corroboration that leucine aminopeptidase in these two species is a hexamer based on two trimers each made of three bilobal promoters.

Refined crystal structures are reported for complexes of Escherichia coli and chicken dihydrofolate reductase containing the antibiotic trimethoprim (TMP). Structural comparison of these two complexes reveals major geometrical differences in TMP binding that may be important in understanding the stereo-chemical basis of this inhibitor's selectivity for bacterial dihydrofolate reductases. For TMP bound to chicken dihydrofolate reductase we observe an altered binding geometry in which the 2,4-diaminopyrimidine occupies a position in closer proximity (by approximately 1 A) to helix alpha B compared to the pyrimidine position for TMP or methotrexate bound to E. coli dihydrofolate reductase. One important consequence of this deeper insertion of the pyrimidine into the active site of chicken dihydrofolate reductase is the loss of a potential hydrogen bond that would otherwise form between the carbonyl oxygen of Val-115 and the inhibitor's 4-amino group. In addition, for TMP bound to E. coli dihydrofolate reductase, the inhibitor's benzyl side chain is positioned low in the active-site pocket pointing down toward the nicotinamide-binding site, whereas, in chicken dihydrofolate reductase, the benzyl group is accommodated in a side channel running upward and away from the cofactor. As a result, the torsion angles about the C5-C7 and C7-C1' bonds for TMP bound to the bacterial reductase (177 degrees, 76 degrees) differ significantly from the corresponding angles for TMP bound to chicken dihydrofolate reductase (-85 degrees, 102 degrees). Finally, when TMP binds to the chicken holoenzyme, the Tyr-31 side chain undergoes a large conformational change (average movement is 5.4 A for all atoms beyond C beta), rotating down into a new position where it hydrogen bonds via an intervening water molecule to the backbone carbonyl oxygen of Trp-24.

X-ray structural results are reported for 10 triazine and pyrimidine inhibitors of dihydrofolate reductase, each one studied as a ternary complex with NADPH and chicken dihydrofolate reductase. Analysis of these data and comparison with structural results from the preceding paper (Matthews, D.A., Bolin, J.T., Burridge, J.M., Filman, D.J., Volz, K.W., Kaufman, B. T., Beddell, C.R., Champness, J.N., Stammers, D.K., and Kraut, J. (1985) J. Biol. Chem. 260, 381-391) in which we contrasted binding of the antibiotic trimethoprim (TMP) to chicken dihydrofolate reductase on the one hand with its binding to Escherichia coli dihydrofolate reductase on the other, permit identification of differences that are important in accounting for TMP's selectivity. The crystallographic evidence strongly suggests that loss of a potential hydrogen bond between the 4-amino group of TMP and the backbone carbonyl of Val-115 when TMP binds to chicken dihydrofolate reductase but not when it binds to the E. coli reductase is the major factor responsible for this drug's more potent inhibition of bacterial dihydrofolate reductase. A key finding of the current study which is important in understanding why TMP binds differently to chicken and E. coli dihydrofolate reductases is that residues on opposite sides of the active-site cleft in chicken dihydrofolate reductase are about 1.5-2.0 A further apart than are structurally equivalent residues in the E. coli enzyme.

In an x-ray diffraction study by the isomorphous replacement method, the structure of the complex of aspartate carbamoyltransferase (EC 2.1.3.2) bound to the bisubstrate analogue N-(phosphonacetyl)-L-aspartate has been solved to 2.9-A resolution (R = 0.24). The large quaternary structural changes previously deduced by molecular replacement methods have been confirmed: the two catalytic trimers (c3) move apart by 12 A and mutually reorient by 10 degrees, and the regulatory dimers (r2) reorient each about its twofold axis by about 15 degrees. In this, the T-to-R transition, new polar interactions develop between equatorial domains of c chains and the Zn domain of r chains. Within the c chain the two domains, one binding the phosphonate moiety (polar) and the other binding the aspartate moiety (equatorial) of the inhibitor N-(phosphonacetyl)-L-aspartate, move closer together. The Lys-84 loop makes a large relocation so that this residue and Ser-80 bind to the inhibitor of an adjacent catalytic chain within c3. A very large change in tertiary structure brings the 230-245 loop nearer the active site, allowing Arg-229 and Gln-231 to bind to the inhibitor. His-134 is close to the carbonyl group of the inhibitor, and Ser-52 is adjacent to its phosphonate group. However, no evidence exists in the literature for phosphorylation of serine in the mechanism. Residues studied by other methods, including Cys-47, Tyr-165, Lys-232, and Tyr-240, are too far from the inhibitor to have a direct interaction.